Skip to main content

Full text of "The origin of floral structures through insect and other agencies"

See other formats

G. N. Jonts 

Professor of Botany 

University oi Ulinota 






Each book complete in One Volume, 12mo, and bound in Cloth, 

1. FORMS OF WATER : A Familiar Exposition of the Origin and Phenomena 

of Glaciers. By J. Ttndall, LL. D., F. R. S. With 25 Illustrations. $1.50. 

2. PHYSICS AND POLITICS ; Or, Thoughts on the Application of the Prin- 

ciples of "Natural Selection" and "Inheritance" to Political Society. 
By WAI.TER Bagehot. $1.50. 

3. FOODS. By Edwakd Smith, M. D., LL. B., F. R. S. With numerous Illus- 

trations. $1.75. 

4. MIND AND BODY : The Theories of their Relation. By Alexander Bain, 

LL. D. With 4 Illustrations. $1,50. 

5. THE STUDY OF SOCIOLOGY. By Herbert Spencer. $1.50. 

6. THE NEW CHE^NnSTRY. By Professor J. P. Cooke, of Harvard Univer- 

sity. With 31 Illustrations. $2.00. 

7. ON THE CONSERVATION OF ENERGY. By Balfour Stewart, M. A., 

LL.D., F.R.S. With 14 Illustrations. $1.50. 

8. ANIMAL LOCOMOTION ; or. Walking, Swimming, and Flying. By J. B. 

Pettigrew, M. D., F. R. S., etc. With 130 Illustrations. $1.75. 



10. THE SCIENCE OF LAW. By Professor Sheldon Amos. $1.75. 

11. ANIMAL MECHANISM: A Treatise on Terrestrial and Aerial Locomotion. 

By Professor E. J. Marey. With 117 Illustrations. $1.75. 


ENCE. By J. W. Draper, M. D., LL. D. $1.75. 


Schmidt (Strasburg University). With 26 Illustrations. $1.50. 


Hermann Vogel (Polytechnic Academy of Berlin). Translation thoroughly 
revised. With 100 Illustrations. $2.00. 

15. FUNGI : Their Nature, Influences, Uses, etc. By M. C. Cooke, M. A., LL. D. 

Edited by the Rev. M. J. Berkeley, M. A., F. L. S. With 109 Illustrations. 

16. THE LIFE AND GROWTH OF LANGUAGE. By Professor William 

DwiQHT Whitney, of Yale College. $1.50. 

New York . D. APPLETON & CO., 1, 3, & 5 Bond Street. 

TTie International Scientific /Scri^— (Continued.) 


Jevons, M. a., F. R. S. $1.75. 

18. THE NATURE OF LIGHT, with a General Account of PhyBical Optics. By 

Dr. Eugene Lommel. With 188 Llustrations and a Table of Spectra in 
Chromo-lithography. $2.00. 


With 83 Illustrations. $1.50. 

20. FERMENTATION. By Frofessor Schutzenbergeb. With 28 Dlustrations. 


21. THE FIVE SENSES OF MAN. By Profeesor Bernstein. With 91 Illus- 

trations. $1.75. 


fessor PiETRO Blaserna. With numerous Illustrations. $1.50. 

23. STUDIES IN SPECTRUM ANALYSIS. By J. Norman Lockter, F. R. S. 

With 6 Photographic Illustrations of Spectra, and numerous Engravings 
on Wood. $2.50. 


fessor R. H. Thurston. With 163 Illustrations. $2.50. 

25. EDUCATION AS A SCIENCE. By Alexander Bain, LL.D. $1.75. 

26. STUDENTS' TEXT-BOOK OF COLOR; Or, Modem Chromatics. With 

Applications to Art and Industry. By Professor Ogden N. Rood, Colum- 
bia College. New edition. With 130 Illustrations. $2.00. 

27. THE HUMAN SPECIES. By Professor A. de Quatrefages, Membre de 

rinstitut. $2.00. 

28. THE CRAYFISH : An Introduction to the Study of ZoOlogy. By T. H. 

Huxley, F. R. S. With 82 Illustrations. $1.75. 

29. THE ATOMIC THEORY. By Professor A. Wubtz. Translated by E. 

Cleminshaw, F. C. S. $1.50. 


EXISTENCE. By Karl Semper. With 2 Maps and 106 Woodcuts. $2.00. 

31. SIGHT : An Exposition of the Principles of Monocular and Binocular Vision. 

By Joseph Le Conte, LL. D. With 132 Illustrations. $1.50. 


J. Rosenthal. With 75 Uluetrations. $1.50. 

33. ILLUSIONS : A Psychological Study. By James Sully. $1.50. 

34. THE SUN. By C. A. Young, Professor of Astronomy in the College of New 

Jersey. With numerous Illustrations. $2.00. 

New York : D. APPLETON & CO., 1, 3, & 5 Bond Street. 

The International Scientific *Senes— ((.'ontinued.) 

85. VOLCANOES: What they Are and what they Teach. By John W. Judd, 
F.R. S., Professor of Geology in the Royal School of Mines. With 96 Il- 
lustrations, $2.00. 

b6. SUICIDE: An Essay in Comparative Moral Statistics. By Henrt Mor- 
sELLi, M.D., Professor of Psychological Medicine, Royal University, 
Turin. $1.75. 

TION OF WORMS. With Observations on their Habits. By Charles 
Darwin, LL. D., F. R. S. With Illustrations. $1.50. 

Stallo. $1.75. 

39. THE BRAIN AND ITS FUNCTIONS. By J. Ldts. $1.50. 

40. MYTH AND SCIENCE. By Tito Vignoli. $1.50. 

41. DISEASES OP MEMORY: An Essay in the Positive Psychology. By Th. 

RiBOT, author of " Heredity." $1.50. 

42. ANTS, BEES, AND WASPS. A Record of Observations of the Habits of 

the Social Hymenoptera. By Sir John Lubbock, Bart., F. R. S., D. C. L., 
LL. D., etc. $2.00. 

43. SCIENCE OF POLITICS. By Sheldon Amos. $1.75. 

44. ANIMAL INTELLIGENCE. By Geobge J. Romanes. $1.75. 

45. MAN BEFORE METALS. By N. Jolt, Correppondent of the Institute. 

With 148 Illustrations. $1.75. 


fe^^sor in Ordinary of Anatomy at the University of Ziirich. With 47 
Woodcuts. $1.75. 

47. FALLACIES: A View of Logic from the Practical Side. By Alfred 

SiDGWiCK, B. A., Oxon. $1.75. 

48. ORIGIN OP CULTIVATED PLANTS. By Alphonse de Candolle. $2.00. 

49. JELLY-FISH, STAR-FISH, AND SEA-URCHINS. Being a Research on 

Primitive Nervous Systems. By George J. Romanes. $1.75. 


iam KiNGDON Clifford. $1.50. 

51. PHYSICAL EXPRESSION: Its Modes and Principles. By Francis War- 

ner, M. D., Assistant Physician, and Lecturer on Botany to the London 
Hospital, etc. With 51 Illustrations. $1.75. 

52. ANTHROPOID APES. By Robert Hartmann, Professor in the University 

of Berlin. With 63 Hlustrations. $1.75. 

New York : D. APPLETON & CO., 1, 3, & 5 Bond Street. 

Tfie International Scientific Series— (ContinvieA.) 


Oscar Schmidt. §1.50. 

54. COMPARATIVE LITERATURE. By Hutchesok Macattlat Posnett, 

M. A., LL. D., F. L. S., Barrister-at-Law ; Professor of Classics and English 
Literafure, University College, Aukland, New Zealand; author of "The 
Historical Method/' etc. $1.75. 


Professor of Mining and Geology in the Imperial College of Engineering, 
Tokio, Japan. With 38 Figures. $1.75. 

56. MICROBES, FERMENTS, AND MOULDS. By E. L. Tbouessabt. With 

107 Illustrations. $1.50. 


MALS. By Angelo Heilprin. $2.00. 

58. WEATHER. A Popular Exposition of the Nature of Weather Changes from 

Day to Day. With Diagrams. By Hon. Rajlph Abercrombt. $1.75. 

59. ANIMAL MAGNETISM. By Alfred Binet and Charles Fere, Assistant 

Physician at the Salpetriere. $1.50. 

60. INTERNATIONAL LAW, with Materials for a Code of International Law. 

By Leone Levi, Professor of Common Law, King's College. $1.50. 


Sir J. William Dawson, LL. D., F. R. S. $1.75. 

62. ANTHROPOLOGY. An Introduction to the Stndy of Man and Civilization. 

By Edward B. Tylor, D. C. L., F. R. S. Illustrated. $2.00. 


OTHER AGENCIES. By the Rev. George Henslow, M. A., F. L. S., 
F. G. S . With 88 Hlustrations. 

















The belief that we must look mainly to the environment 
as furnishing the influences which induce plants to vary 
in response to them — whereby adaptive morphological 
(including anatomical) structures are brought into exist- 
ence — appears to be reviving. To illustrate the progress 
of this belief, I will give a few cases. 

In 1795, Geoffroy Saint Hilaire "seems to have 
relied chiefly on the conditions of life, or the ' monde 
ambiant,' as the cause of change." * 

In 1801, Lamarck " attributed something to the direct 
action of the physical conditions of life " as the means 
of modification, "something to the crossing of already 
existinsT forms, and much to use and disuse." 

In 1831, Mr. Patrick Matthew (who, like Dr. W. C. 
"Wells in 1818, anticipated Mr. Darwin in the theory of 
"natural selection") "seems to have attributed much 
influence to the direct action of the conditions of life." 

* I quote from Mr. Darwin's "Historical Sketch" in his Origin of 
Species, 6th ed., 1878. 


In 1844, the " Vestiges of Creation " appeared. The 
author suggests that "impulses" were imparted to the 
forms of life, on the one hand advancing them, and on 
the other hand tending to modify organic structures in 
accordance with external circumstances ; the effects thus 
produced by the conditions of life being gradual. 

In 1852, Mr. Herbert Spencer " attributed the modifi- 
cations [of species] to the change of circumstances." 

In 1859, " The Origin of Species " appeared. Mr. 
Darwin did not at first seem to lay so much stress as 
his predecessors upon the action of the environment as 
a cause, for he says : " It is curious how largely my 
grandfather, Dr. Erasmus Darwin, anticipated the views 
and erroneous grounds of opinion of Lamarck." Again, 
in speaking of the constancy of some varieties, he says, 
" Such considerations incline me to lay less weight on 
the direct action of the surrounding conditions, than on 
a tendency to vary, due to causes of which we are quite 
ignorant."* He had, however, previously said, " Changed 
conditions of life are of the highest importance in 
causing variability. ... It is not probable that vari- 
ability is an inherent and necessary contingent under 
all circumstances." f 

With regard to my own opinion, having been early 
and greatly interested in Paley's "Natural Theology," 
as well as the " Vestiges " when Mr. Darwin's work 

* Or. of Sp., p. 107. t Ihid., p. 31. See also Desc. of Man, ii., p. 388. 


appeared, the great difficulties I felt in accepting 
natural selection as any real origin of species lay, 
first, in the seeming impossibility of the histological 
minutise of the organs in adaptation having been selected 
together; and, secondly, in the idea that all those 
wonderful and "purposeful" structures which Paley 
thought could only have been " designed," could be the 
ultimate result of any number of accidental and appa- 
rently at first ''purposeless" variations. In a broad 
sense natural selection seemed obviously true ; for 
Geology had revealed the fact that the world had been 
peopled over and over again by old forms dying out and 
new forms coming in ; so that although it might account 
for the extinction of the former, it did not seem to me 
capable to account for the origin of the latter. I, there- 
fore, still looked to the environment as affording a better 
clue to the source of variations.* 

In 1869, when watching a large humble-bee hanging 
on to the dependent stamens of Epilohiumi angusti- 
folium, the idea first occurred to me that insects them- 
selves might be the real cause of many peculiarities in 
the structure of flowers. The thought passed through 
my mind that the way the stamens hung down might 
perhaps have become an hereditary effect from the 
repeatedly applied weight of the bees. 

In 1877, I advanced this idea as a speculation 

* See Letter to Nature, vol. v., p. 123. 


when suggesting the origin of nectaries and irregu- 
larities of flowers in my paper on " Self-fertilisation of 
Flowers." * 

In 1880, Mr. A. R. Wallace reviewed Dr. Aug. Weis- 
mann's " Studies in the Theory of Descent." f In this 
work the author says : " According to my view, trans- 
mutation by purely internal causes is not to be enter- 
tained. . . . The action of external inciting causes is 
alone able to produce modifications." Mr. Wallace adds 
that he had "arrived at almost exactly similar con- 

In 1881, when reviewing Paul Janet's work on 
" Final Causes," J I took occasion to remark that " I 
regarded the environment as by far the most important 
"cause" of variations, in that it influences the organ- 
ism, which, by its inherent but latent power to vary, 
responds to the external stimulus, and then varies 

In 1881, appeared the first really systematic treatise 
that I know of, by Dr. C. Semper, § which dealt with the 
oriorin of variations in animals as beino^ referable to the 

In 1884, Dr. A. de Bary's "Comparative Anatomy of 
the Vegetative Organs of the Phanerogams and Ferns," 

* Trans. Lin. Soc, 2nd ser., Bot., vol. i., p. 317. 
t Nature, xxii., p. 141. X Modern Review, 1881, p. 53. 

§ The Natural Conditions of Existence as they affect AnimalSy" Intern. 
Sci. Ser., vol. xxxi. 


was published in English. In the Introduction, the 
author writes as if it were a perfectly well understood 
thing that species have arisen by adaptations to the 
influences of the environment.* 

In 1886, Mr. Herbert Spencer contributed two articles 
on " The Factors of Organic Evolution " to the Nineteenth 
Century. \ In these he showed, from many passages in 
Mr. Darwin's works, especially "Animals and Plants 
under Domestication" and in his later volumes, that he 
became much more favourably inclined to the belief that 
the effects of the environment were accumulative, and 
that in the course of some generations the variations set 
up tended to cease and become fixed. Mr. Spencer par- 
ticularly notes the change of view, as illustrated by the 
expression " little doubt " being replaced by " no doubt " 
in the following sentence : " I think there can be no 
doubt that use in our domestic animals has streno^thened 
and enlarged certain parts, and disuse diminished them ; 
and that such modifications are inherited." + It may 
be added that in " The Cross and Self Fertilisation of 
Flowers" (1876), and in "Forms of Flowers" (1877), 
Mr. Darwin makes many observations upon the effects 
of the external conditions upon plants as influencing and 
modifying them in various ways. It is curious to note 
that the three influences upon which Lamarck laid 

* See, e.g., p. 25. f See p. 570 and p. 749. 

X "Use" and "disuse" in animals corresponds to what I have 
called " hypertrophy " and " atrophy " in plants, in this work. 


emphasis are just those which Mr. Darwin himself 
latterly, though often indirectly perhaps, laid stress upon 
in his experiments, viz. crossing, use and disuse, and the 
physical conditions of life. 

In 1886, also appeared an article in Nature, entitled, 
" Plants considered in Relation to their Environment." 
It was not signed, but the author alludes to the external 
conditions as bringing about all sorts of changes in the 
vegetative system. He stops short of discussing floral 

In 1886, Dr. Vines' " Physiology of Plants" appeared. 
After discussing various views and theories of reproduc- 
tion, he observes, that " variability was first induced as 
the response of the organism to changes in the conditions 
of life." * . . . We conclude, then, that the production of 
varieties is the result of the influence of the conditions 
of life, t 

In the last page of his work. Dr. Vines calls attention 
to Naegeli's view as follows : " Naegeli suggests, and his 
suggestion is worthy of serious consideration, that there 
is an inherent tendency to a higher organisation, so that 
each succeeding generation represents an advance, ... as 
in cases of what is termed saltatory evolution." Thus, 

* Page 676. Dr. Vines here uses almost identically the same words 
as myself in 1881. I have just found that Mr. St. G. Mivart said much 
the same in 1870, Genesis of Species, p. 269. See also O. Schmidt's 
Doctrine of Descent and Darwinism, p. 175. 

t Page 679. 


while Mr. Darwin seems at last to have tacitly accepted 
Lamarck's ideas, at least to a considerable extent, we 
have here a return in 1887 to the views of the author of 
the " Vestiges " of 1884?. 

1888. I have attempted in the present work to return 
to 1795, and to revive the "Monde ambiant" of Geoffroy 
Saint Hilaire, as the primal cause of change. My object 
is to endeavour to refer every part of the structures of 
flowers to some one or more definite causes arising from 
the environment taken in its widest sense. To some 
extent the attempt must be regarded as speculative ; 
and, therefore, any deductive or a jpriori reasonings met 
with must be considered by the reader as being suggestive 




I. General Principles ... ... .- ••• 1 

11. The Principle op Number ... ... ... ^ 

III. The Principle of Number — Continued ... ... 25 

IV. The Principle of Arrangement ... ... 39 

V. The Principle of Cohesion ... ... ... 48 

VI. The Principle of Cohesion— ContmMed! ... 54 

VII. The Principle of Cohesion— CojiftnMe(i ... ... 62 

VIII. The Principle of Adhesion ... ... ... 78 

IX. The Cause of Unions ... ... ... ... 84 

X. The Receptacular Tube ... ... ... 89 

XL The Forms OF Floral Organs ... ... ... 101 

XII. The Origin of " Zygomorphism " ... ... 116 

Xm. The Effects of Strains on Structures ... ... 123 


XV. The Origin of Floral Appendages ... ... 133 

XVI. Secretive Tissues ... ... ... ••• 140 

XVII. Sensitiveness and Irritability of Plant Organs ... 151 

XVIII. Origin of Conducting Tissues ... ... 164 




XIX. Colours of Flowers 

XX. The Emergence of the Floral Whorls 

XXI. The Development of the Floral Whorls 

XXII. Heterogamy and Autogamy ... 

XXIII. Heterostylism 

XXIV. Partial Diclinism 

XXV. Sexuality and the Environment... 

XXVI. Degeneracy of Flowers 

XXVII. Degeneracy of Flov^-ees— Continued 

XXVIII. Progressive Metamorphoses ... 

XXIX. Eetrogressive Metamorphoses 

XXX. Phyllgdy of the Floral Whorls 

XXXI. The Varieties of Fertilisation ... 







1. DiagT-am of a typical flower composed of six whorls ... 3 

2. Diagram of the positions of opposite leaves, illustrating the 

method of passage to alternate arrangements ... ... 11 

3. Diagrams of floral sestivations, showing the passage from the 

two-fifth or quincuncial, to the contorted ... ... ... 15 

4. Diagram of flower of Garidella, with stamens superposed to 

petals ... ... ... ... ... ... ... ... 21 

5. Diagram of flower of Helleborus niger with stamens super- 

posed to twenty-one nectaries 22 

6. Diagrams illustrating the anatomy of the floral receptacle of 

a Wallflower, showing the origin of the floral members ... 32 

7. Diagram of the leaf -traces in the stem of Arahis alhida ... 39 

8. Vertical and transverse sections of the wall of the inferior 

ovary of Canvpanula medium, showing how the sepaline 
cords originate those of the rest of the floral organs (see 
fig. 15, p. 71) ' ... ... 43 

9. Flower of Phyteuma, showing cohesion by contact and con- 

genital, in the corolla ... ... ... ... ... ... 50 

10. Flower of IfimwZws undergoing " dialysis " ... ... ... 51 

11. Stamens of Centaurea, showing syngenesious anthers ; method 

of fertilisation by " piston-action " (6), nectary and direc- 
tion of insect-proboscis, etc. ... ... ... ... ... 60 

12. Anatomy of the floral receptacle of Hellebore, showing the 

changes in orientation of the cords ... ... ... ... 64 

13. Anatomy of the floral receptacle of Pelargonium, showing 

changes in the orientation, in the separation and in the 
imion of the cords ... ... ... ... ... «.. G^ 



14. Anatomy of the floral receptacle of Ivy, showing the multi- 

plication and differentiation of the cords, etc. ... ... 68 

15. Anatomy of the floral receptacle of Campanula medium^ 

showing the distribution of the cords, etc. (see fig. 8, p. 43) 71 

16. Origin and development of the ovule in Bef a ... ... ... 73 

17. Carpels of Acer, showing the thickened bases, preparatory for 

ovules 75 

18. A separate carpel of Primula sinensis, with marginal ovules 

and a " heel-like " process, the origin of the free central 
placenta ... ... ... ... ... ... ... ... 76 

19. Anatomy of the floral receptacles of Lysimachia and Primula, 

showing the cords of five carpels ... ... ... ... 77 

20. Echium, showing declinate stamens and protandrous condition 82 

21. Ovary, stamens, and stigmas of Aristolochia ... ... ... 83 

22. Vertical sections of buds of Pyrus and Cotoneaster, showing 

degrees of adhesion or undifferentiated condition between 

the ovary and receptacular tube ... ... ... ... 90 

23. Orchis Morio (?), with arrest of pistil, the receptacular tube 

represented by a rod-like pedicel. Two anthers are deve- 
loped instead of one (a) ... ... ... ... ... 92 

24. Receptacular tube of Rose, bearing a leaf and a stipular sepal 93 

25. Vertical section of the receptacular tube of Hawthorn, with 

supernumerary carpels arising from the summit ... ... 93 

26. Leaves of Pear with hypertrophied and sub-fasciate petioles 94 

27. Fuchsia with foliaceous sepals, partly detached from the ovary 94 

28. Anatomy of the receptacular tube of Prunus, showing the 

origin of the petaline and staminal cords ... ... ... 95 

29. Part of the receptacular tube of Cherry, showing the distri- 

bution of cords in the sepaline lobes 97 

30. Anatomy of inferior ovary of Alstrcemeria, showing the junc- 

tion between the ovary and the tube ... ... ... 97 

31. Flower of Duvernoia, showing its adaptability for intercrossing 107 

32. Flower of Calceolaria, showing thickened ridges, etc., and 

adaptability for intercrossing ... ... ... ... ... 109 

33. Flower of Bictamnus, showing declinate stamens and displace- 

ment of petals 110 

34. Flower of Epilohium angustifolium, showing dependent sta- 

mens and displacement of petals ... ... ... ... Ill 



35. Flower of Veronica Chamcedrys, showing method of fertilisa- 

tion by insects, and degeneracy of anterior petal ... ... Ill 

36. Flower of Teucrium, to show effect of weight of insect with 

exposure of stamens ... ... ... ... ... ... 117 

37. Diagrams of Narcissus cernuus^ to show instability in the 

heterostylism and lengths of stamens ... ... ... 121 

38. Basal end of a Pear, to show cause of thickening in response 

to forces 124 

39. Diagram of a declinate bough, showing distribution of forces 125 
40 a. A diagram of declinate stamens, to show distribution of forces 126 
40 &. Flower of LamiMwi aZbwm, to show distribution of forces ... 126 

41. Base of flower of Amaryllis, showing the honey-protector ... 134 

42. Adhesive epidermal cells of roots of Orchids ... ... ... 137 

43. Stipules of 7mpa^^e?ls, showing nectariferous tissue ... ... 140 

44. Petals passing into nectariferous stamens of Atragene ... 141 

45. Cells of hair of TradescarUia, showing the state of protoplasm 

before and after excitation by electricity ... ... ... 152 

46. Climbing peduncle of Uncaria, thickened after irritation by 

the support ... ... ... ... ... ... ... 156 

47. Flower of Genista tinctoria, before and after mechanical 

irritation ; the claws of the keel and wing petals being 

in unstable equilibrium ... ... ... ... ... 160 

48. Flowers of Lopezia in three stages, showing movements of the 

staminode and stamen ... ... ... ... ... 161 

49. Flower of Medicago saliva, before and after mechanical irrita- 

tion, the staminal tube being in unstable equilibrium . . . 162 

50. Transverse sections of conducting tissues of Fumaria, Ruhus, 

and of a Crucifer ... ... ... ... ... ... 164 

51. Diagram of emergence of the petaline stamens of Peganum 

outside the sepaline ... ... ... ... ... ... 189 

52. Flower-bnd, and same opened, of Stellaria media, showing 

conditions of degeneracy and adaptations for self-fertili- 
sation ... ... ... ... ... ... ... ... 255 

53. Flower-bud, and essential organs of Epilohium montanum, 

showing positions for self -fertilisation ... ... ... 255 

54. Styles and stigmas of the two forms of Pansy, showing the 

conditions which (a) prevent and (6) secure self-ferti- 
lisation, respectively ... ... ... ... ... ... 255 



55. Styles and stigmas of self -fertilising forms of Pansy ... ... 257 

56. Details of structure of cleistogamous Violets ... ... ... 258 

57- Details of structure of cleistogamous Oxalis Acetosella ... 260 

58. Flower.bud and stamens of cleistogamous Impafiens ... ... 261 

59. Flower-bud and section of cleistogamous Lamium amplexi- 

caule ... ... ... ... ... ... ... ... 261 

60. Corolla, stamens, and style of Salvia clandestina, showing 

adaptations for self -fertilisation ... ... ... ... 262 

61. Transitional forms between bracts and leaves of Hellehorus 

viridis ... ... ... ... ... ... ... ... 286 

62. Inflorescence of Cornus florida, showing floral mimicry ... 287 

63. Inflorescence of Dariumia, showing floral mimicry ;.. ... 287 

64. Involucral bract of Nigella, bearing an anther 288 

65. Glumes of Lolium, both antheriferous and stigmatiferous ... 288 

66. EammcMZws with a foliaceous sepal ... ... ... ... 289 

67. Foliaceous calyx of TrifoUum repens with stipulate leaves, 

borne by the receptacular tube ... ... ... ... 289 

68. Flower and leaf of llitsscewda 290 

69. Linai-ia with one sepal petaloid ... ... ... ... ... 291 

70. Calyx of Garden Pea with carpellary lobes 292 

71. Ovuliferous sepal of "Violet ... ... ... ... ... 292 

72. Corolla of Foxglove with filamentous processes, some being 

antheriferous ... ... ... ... ... ... ... 292 

73. J-gMiZegrmj the corolla with polleniferous spurs ... ... 293 

74. Ovuliferous petals, etc., of Begronm ... ... ... ... 293 

75. Ovuliferous anthers of Sempervivum 294 

76. Stigmatiferous and ovuliferous stamens of £egroma ... ... 294 

77. Carpels and ovules originating from a placenta of Carnation, 

the carpels again ovuliferous (a) ... ... ... ... 295 

78. Stameniferous carpels of Willow and Ranunculus auricomus 296 

79. Petaliferous placentas of Cardamine pratensis and of Rhodo- 

dendron ... ... ... ... ... ... ... ... 296 

80. Metamorphosed sub-petaloid carpel of Polyanthus ... ... 297 

81. Foliaceous connective of Petunia . . ... ... ... 298 

82. Petalody, or " hose-in-hose " fonn, of connective in a double 

Columbine (Aquilegia) ... ... ... ... ... ... 298 

83. Foliaceous stamen and petal of the Alpine Strawberry and 

stamen of the Green Rose ... ... ... ... ... 302 



84. Stamen of Jatropha Pohliana, with foliaceous membraTies to 

the anther-cells 302 

85. Metamorphosed and foliaceous ovules of Mignonette ... ... 305 

86. Metamorphosed and foliaceous ovules of Sisijmhrium Alliaria 306 

87. Tubular excrescence on the labellum of Cattleya, homologous 

with an ovule ... ... ... ... ... ... ... 306 

88. Multifold carpels, with ovuliferous margins, from a malformed 

Primrose 308 






Introductory. — Mncli has been written on the structure of 
flowers, and it might seem almost superfluous to attempt to 
say anything more on the subject ; but it is only within the 
last few years that a new literature has sprung up, in which 
the authors have described their observations and given 
their interpretations of the uses of floral mechanisms, more 
especially in connection with the processes of fertilisation. 

Moreover, there is a considerable amount of scattered 
literatnre on special points which seems never to have been 
collated, so as to show the reletive significance of the dif- 
ferent classes of observations to which the authors have 
devoted themselves respectively. The consequence is, that, 
good as each in itself may be, it often requires the help of 
other classes of facts to enable one to fully elucidate any 
question to be discussed. 

Now, the primary object of the first really scientific study 
of plants was their classification, and no longer with the 
sole view of ascertaining the real or imaginary medicinal 
uses of herbs ; as had been the case in Gerarde's time, when 
a botanist and a herbalist were one and the same. 


Systematic botanists, however, have hitherto invariably 
contented themselves with observing differences of structure 
only; and paid little or no attention to the "why" and the 
" wherefore " of the differences they seized upon as being 
more or less important for the purpose of distinguishing 
species. When, however, the desirability of a more thorough 
knowledge of the origin of parts of plants as interpreting mor- 
phological characters was felt, developmental history began 
to be studied ; a method strongly insisted upon by Schleiden, 
for example ; and the most elaborate result of this method of 
investigation is undoubtedly Payer's Trait e d' Org anogenie Com- 
parSe de la Fleur, published in 1857: but if it be thought 
sufficient to limit the study of flowers to tracing their mor- 
phological development alone, one soon begins to see that it 
is far from being so, and, taken by itself, it may lead one into 
false interpretations, so that to the study of development 
must be added that of anatomy. To Ph. van Tieghem we 
are indebted for an elaborate treatise, entitled RecJierclies sur 
la Structure du Pifitil et sur V Anatomee Coniparee de la Fleur 
(1871), dealing with the more minute details of floral struc- 
tures. This treatise, however, still leaves much to be 

Besides these methods, analogy and especially teratology 
furnish assistance of no mean value. Here we are especially 
indebted to Dr. M. T. Masters for his standard work on 
Teratology ,^ 

Now, any one of these methods taken alone would be 
insufficient, and in many cases would be far from thoroughly 
accounting for particular points under consideration. 

Hence to arrive at a complete interpretation of the origin 
of every sort of structure to be found in floAvers, it can only 

* A Gerinan edition, Pflanzen Teratologic, ed. Tainmer, 1886, has 
Euraerous additions. 



be done by calling in the aid of each and all tliese methods 
to the very utmost extent possible. 

Lastly, to attempt any theoretical exposition of the evo- 
lutionary history of flowers, considerable caution is required ; 
for the causes of variation are generally so obscure, the 
chances of seeing them in activity so small, and experimental 
methods of verification well-nigh impossible, that specula- 
tions on this subject cannot altogether escape the bounds of 
hypothesis so as to become demonstrable facts. Hence 
observations Avhich I shall make later on, with reference to 
the origin of existing floral structures, will not profess to be 
anything more than theoretical, and at most only a "work- 
ing hypothesis " for future investigations. 

The Structure of a Typical Flower. — Before consider- 
ing how the innumerable forms 
of flowers deviate from one 
another, it is advisable to assume 
some typical form or plan as a 
preliminary basis to start from, 
or to which all flowers, if pos- 
sible, may be referred as a 
standard. It would be quite 
possible to adopt some kind of 
flower as it exists in nature, 
but as this would be arbitrary, 
it may be better to take an ideal 
type, and the diagram (Fig. 1) will answer the purpose, 
in which the outermost circle is supposed to represent a 
cross section of the five Sepals constituting the Calyx. The 
second circle is that of the five Petals of the Corolla. The 
third stands for the Anthers of the five Stamens superposed 
to the sepals ; the fourth being those of five Stamens super- 
posed to the petals. These two whorls of stamens together 

Fig. 1. -Diagram of a typical flower. 


constitute the Andrcecium. Lastly, there are represented 
two * whorls of Carpels forming the Gjnoecium j or Pistil. 
The outermost whorl of carpels is superposed to the sepals, 
the innermost to the petals. 

There may be additional structures in flowers, such as 
disks, honey-glands, etc. ; but as these, when they occur on 
the floral-receptacle, are merely cellular protuberances and 
form no part of the floral whorls proper — not being foliar in 
their origin — they may be omitted, especially as their posi- 
tion is by no means constantly the same in all flowers. J 

The Principles of Variation. — Having thus assumed an 
ideal type, we may at once consider the " Principles of 
Variation," as I propose to call them, in accordance with 
which the different members of flowers can be altered; so 
that by means of various combinations of these principles all 
the flowers in the Vegetable Kingdom can be brought under 
this one fundamental plan. 

There are five principles which require special considera- 
tion. They are usually designated by the terms Number, 
Arrangement, Cohesion, Adhesion, and Form. 

"Number" refers to the number of whorls and the 
number of parts in each whorl. If two or more whorls 
contain the same number of parts or be multiples of one 
another, they are said to be " symmetrical " or " isomerous." 
If they differ in the number of parts they are " unsym- 
metrical " or " anisomerous." 

"Arrangement" refers to the relative positions of the 

* Why I assume tivo whorls for the pistil, instead of one only, as is 
generally done, will be understood hereafter. I have since found that 
Robert Bi'own came to the same conclusion {Col. Woi-Jcs, i. 293). 

f I adopt the spelling Gynoecium for the sake of uniformity ; it may 
be regarded as a shortened form of Gyncecoscium. 

;J: I do not here allude to certain glandular structures, which may 
be the homologues of arrested organs. 


different whorls, as well as of those of the individual members 
of the whorls with regard to each other. 

" Cohesion " signifies the union of parts of any, but of the 
same whorl. The original or ancestral condition of the 
parts composing every whorl is presumed, on the principles 
of evolution, to have been one of entire freedom; so that 
the members were as completely separate or free as, for 
example, tbey are in a Buttercup. Reversions to this con- 
dition of freedom may occur, and then the process is called 
" dialysis , " as in the case of a polypetalous Campanula 
occasionally cultivated as a garden plant. 

" Adhesion " signifies the union of parts of different 
whorls ; as well as that between the ovary and the recepta- 
cular tube, constituting the so-called inferior ovary. I regard 
adhesion as representing a more advanced or a more highly 
differentiated state than that of cohesion. Reversions may 
occur by " solution," which brings about a freedom of parts 
normally united, as in the abnormal cases of Apples, double 
Saxifrage, members of the UmhelUferce, etc., which have all 
their parts perfectly free, though with inferior ovaries under 
ordinary circumstances. 

" Form " refers to the shape of the organs ; such as those 
of sepals and petals upon which generic characters are so often 
founded, the length of the filaments, and other peculiarities. 
If all the parts of any whorl be exactly alike, it is said to 
be " regular ; " if not, the whorl will be " irregular." 

The above five principles constitute the most important 
in accordance with which Nature has brought about the 
infinite diversity which exists in the Floral world. There are 
minor distinctions, hereafter to be considered, such as colours, 
scents, etc. ; but they are of less importance in investigating 
the causes at work which have evolved specific and generic 
differences amongst flowering plants. 


There is anotlier point which may be here noticed. That 
a flower-bud is a metamorphosed leaf-bud is now an accepted 
fact ; but an obvious difference between them consists in the 
arrested state of the axis of the former, constituting the floral 
receptacle ; and the question arises, how has this arrest been 
brought about ? Like all other peculiarities of structure 
to be described, I would attribute the aiTest primarily to the 
altered nature of the foliar organs on becoming members 
of flowers. Thus, a Fir-cone and a Buttercup are arrested 
branches ; but when the parts of a flower are reduced in 
number, and instead of being in a continuous spiral are 
grouped in " compressed cycles," * I would then (hypo- 
thetically) attribute this further reduction of the axis, as 
well as other features hereafter to be described, to the 
irritation of insects in probing for juices, and causing 
nectaries to be formed, f It is the commonest thing for leaf- 
buds to be arrested, and sometimes metamorphosed as well, 
by insects puncturing and depositing their eggs in them. 
Such may be seen on the terminal shoots of Yews, Thyme, 
and in certain kinds of Oak-galls, etc. In all such cases the 
immediate effect is the total arrest of the axis, though the 
leaves may be but slightly altered, as in the Yew. How 
the various metamorphoses of leaves into petals, etc, has 
followed will be discussed later on. 

It must not be forgotten, however, that the tendency to 
shorten the axis is primarily, in some cases, due to the 
altered structure of the foliar organs, as in Gymnosperms ; 
whereby they undertake the reproductive functions. At 
the same time, I think insects have had a good deal to do 
with it, in many other phanerogams, which have but few 
parts to their whorls. 

Each of the above principles must now be considered in 

* See pp. 41, 42. f See p. 140, seqq. 



Number — General Observations. — The first principle of 
Variation to be considered is that of the number of parts 
composing the different whorls of flowers. There are good 
reasons for considering that six whorls, consisting of five, 
four, three, or two parts each, as the case may be, should be 
regarded as the theoretically complete number of verticils 
of any flower. 

Anatomical investigations prove that the rale is for the 
pedicel to contain — at least, immediately below the flower, — 
if the latter be pentamerous, ten more or less distinct fibro- 
vascular cords, five of which belong to the sepals and five to 
the petals ; if it be hexamerous, there will be six cords, three 
for each whorl of the perianth. Each of these cords can 
give rise by branching, first, to a whorl of stamens and 
subsequently to a whorl of carpels, furnishing at least two 
marginal and one dorsal cord for each of the latter. 

In many flowers both whorls of stamens are present, and 
the androecium is then isomerous with the entire perianth. 
More often one whorl is arrested, and then it may be either 
one ; but most usually it is the petaline. On the other 
hand, the calycine may not be developed as in Primroses, 
B'hamnus, etc. 

The absence of the petaline stamens is possibly attribu- 


table to the law of compensation, in consequence of the 
enhanced growth of the corolla, the petals thereby abstract- 
ing the nourishment that would be required by the stamens 
superposed to them. 

That the number of staminal whorls should be two in 
Yerticillate flowers, i.e., equal to the perianth, is apparent 
from the fact that two whorls prevail in Monocotyledons and 
are not at all uncommon in Dicotyledons ; and when the 
petaline whorl alone exists, as in Primulacece and Myrsinece, 
calycine staminodia are sometimes present which tend to 
restore the complete number, as in the genus Samolus in the 
former and in the. tribe Theophrastece of the latter order. 

The reduction of the number of carpels is very generally 
carried to a greater extent than that of the stamens. Assum- 
ing two complete whorls of carpels as the primitive number, 
not only are both rarely to be found in the same flower, as 
in Butomus, but a portion only of one whorl is commoner 
than even a single entire whorl. Thus, two are characteristic 
of Cncciferce, Pohjgalece, and of most of the gamopetalous 
orders ; while one carpel only prevails in Leguminosce and 

That the absence of parts of, as well as of entire whorls 
of flowers as they now exist does not represent primitive 
conditions, is testified to by the frequent occurrence of 
various kinds of degradations, such as were alluded to above in 
the case of the staminodia of Samolus, etc. Thus, with regard 
to the calyx, it is a noticeable fact that when the inflorescence 
consists of a large number of flowers, especially if small and 
closely compacted, there is a strong tendency for the sepals 
to become partially arrested and remain rudimentary, or even 
not to be developed at all. This is particularly observable 
in some epigynous orders as TJmhelliferce^ Araliacece, Cajpri- 
foUacece, Buhiacece, Compositce, etc. 


The degradation of tlie corolla is likewise very common. 
As its enhancement has been due to insect agency, so, 
conversely, its reduction in size, colour, etc., is presumabl}^ 
often the result of the neglect of insects. Consequently 
inconspicuousness becomes a characteristic feature of self- 
fertilising flowers. By increased degradation the corolla may 
disappear entirely, as in Sagina apefala, some cleistogamous 
flowers, and in the Incompletce generally. Sach degradation 
is also characteristic of wind-fertilised flowers. 

As both calyx and corolla maybe degraded and disappear, 
so may the stamens and carpels, unisexual and neuter flowers 
being the result. 

Further observations, however, wdll be made upon this 
subject when treating of the sev^eral whorls respectively, and 
especially when discussing the phenomenon of degeneracy. 

The Origin of Different Numbers. The number of parts 
constituting the floral whorls is, without doubt, primarily 
due to phyllotaxis ; and therefore, to understand why certain 
numbers, such as fives, fours, and threes prevail, it is needful 
to give some preliminary remarks on the principles of leaf 
arrangement. It has long been observed that these are 
referable to two kinds — one in which two or more leaves are 
situated on the same node, when they are decussate,* that 
is to say, each pair or whorl of three or more leaves alternates 
in position with the whorl immediately above and below it. 
The second system is when only one leaf occurs at a node ; 
the leaves are then said to be alternate. The leaves are then 
arranged on a continuous spiral line, and can be represented 
by the fractions of the well-known series i, ^, |, f , -^^^ /y, etc. 
Of these fractions the denominator represents the number of 

* Rare exceptions occur in species of Potamogeton, in which alternate 
intemodes between the distichously arranged leaves are suppressed, so 
that they become opposite, but are all in the same plane. 


leaves in a " cycle," and tlie numerator the number of times 
a spiral line, passing through the position of the leaves, coils 
round the stem in forming a cycle ; thus, with the % arrange- 
ment, any leaf being taken as number 1, the sixth leaf will 
be first that falls in the same vertical line with number l,th6 
leaves 1 to 5 constituting the cycle. The portion of the 
spiral line whicli passes through the leaves 1 to 6 coils ticice 
round the stem, and if projected on a plane forms two circles. 
The angular distance, measured from the centre of the stem 
or circles, between any t^v"o successive leaves is always found 
by multiplying 360° by the fraction : thus f X 360^ = 144°. 

The interpretation, therefore, of the prevailing numbers 
3 and 5 in floral whorls is that they are, in most cases, cycles 
of the ^ or f types respectively ; while 4 is primarily due 
to the union of two pairs of opposite and decussate parts. 

6, 8, 10 are merely the doubles of the preceding, and mostly 
represent two pairs of whorls or cycles blended together, 
thus forming one whorl, or so closely approximated as 
scarcely recognizable as two ; though the rare number 8, in 
some cases, such as Nigella, and Hellehorus foetidus, may repre- 
sent a cycle of the f type. Similarly, the still rarer numbers 

7, 9, and 11 in flowers correspond to the absence of these 
numbers as denominators of any fractions of the above 
prevailing series. 

With the exception of dimerous and tetramerous whorls, 
all the rest are presumably due to alternate arrangements. 
Now, opposite leaves present a more primitive type than 
alternate ; that this is so, is not only reasonable from the 
primordial condition of the cotyledons of Dicotyledons, but 
the transition from an opposite to an alternate condition 
may be often witnessed on rapidly growing stems, such as 
of the Jerusalem Artichoke. Whenever this plant bears 
opposite leaves below, and alternate leaves above, it will be 


found that the an-augement of the latter is almost invariably 
represented by the | type. It is secured by developing inter- 
nodes between the two opposite leaves of each pair, and by 
shifting their positions so as to acquire ultimately an angular 
divergence of 144°.* 

The feature to be especially observed in the transitions 
from opposite to alternate arrangements is the order in which 
the opposite leaves separate so as to assume successive 
positions on the continuous spiral line passing through their 
insertions, when they have become alternate. This will be 
understood from the accompanying diagram, in which the 
numbers represent the order which the leaves will ultimately 
assume on the f type ; tlioagh they ai'e placed as if still 
opposite and decussate. The numbers 1 and 2, 3 and 4, 5 
and 6, etc., represent the successive pairs of opposite leaves, 
the arrows showing the direction of the spiral. 

It will be at once ob- 
served that the numbers 6, 
9, 14, and 22 are in the 
same row, and correspond to 
the divergences |, |, -/g, 
^V No. 17 falls into the 
series |, and completes the 
second cycle of that type 
from No. 1. 

It may be observed 
here, as occasion will arise 
for a fuller allusion to the 
significance of the fact, that, with the sole exception of the 

* I have fully explained this in my paper, On the Variations of the 
Angular Divergences of the Leaves of Helianthns Tuberosus, Trans. Lin. 
Soc, vol. xxvi,, p. 647. See also On the Origin of the Prevailing Systems 
of Phyllotaxis, I.e., 2nd series, vol. i. p. 37. 


distichous or ^ type, every other arrangement always has 
ih7'ee leaves in every projected circle. 

It may be noticed that No. 4 not only does not occur in 
the row 1, 6, 9, etc., — a fact which corresponds with the rarity 
of a ternary arrangement occurring amongst flowers of 
Dicotyledons, — but in order to fall over No. 1 it would have 
to pass through 270°, that is from right to left, practically 
an impossibihty; so that Avhen "threes" are met with in 
Dicotyledons we must look for some other interpretation 
than to refer them to the ^ type. 

The numbers 7 and 11, as stated, are extremely rare in 
flowers, and this is in accordance with the fact that they 
belong to another series, viz. ^, |, |, y\, y\, etc , which is rarely 
represented in natjire. Examples, however, will be found in 
the leaves of Sedum reflexum, on some branches of Araucaria 
imbricata, and sometimes in the Jerusalem Artichoke. In 
the last case, it will be discovered that the heptastichous or 
I" type arises out of verticils of threes, in precisely the same 
way as the pentastichous or f type does from an opposite and 
decussate arrangement ; and as there are always four leaves 
in every projected circle, for every type of this series, except- 
ing the first or ^, it can only occur where the leaves are 
narrow or are short, or do not occupy too much space so as 
to overshadow one another. 

Variations in the Floral Symmetry. — Besides the fact 
that certain numbers are often characteristic of certain 
species, genera, or even orders, great variations in the sym- 
metry exist, not only in different genera of the same order, 
but in different species of the same genus.* 

Now, with reference to this latter fact, it must be borne 
in mind that flowers are so highly differentiated from the 

* See note by the author, On the Causes of the Numerical Increase of 
Parts of Plants, Journ. Lin. See. Bot., xvi. p. 1. 


leaf type, that they hav^e undergone such wonderful transfor- 
mations and adaptations to insect and other agencies and to 
their environing conditions, so that the simple and original 
laws governing the arrangement of the leaves, here pro- 
pounded for the origin of what may be called the " primitive 
symmetry" of the floral organs, have become in many cases, 
masked or interfered with. Hence, to deduce those original 
laws from the present structure of flowers, it is not only neces- 
sary to consider the floral symmetry of an immense number 
of genera, and so ascertain what are the relative proportions 
of certain numbers when associated with alternate and oppo- 
site leaves respectively, but to discover what may have been 
the interfering causes which have modified what would 
have been the immediate effects of the fundamental laws of 

Thus, it will be found that the numbers of the parts 
of whorls are liable to vary on their own account, while 
the arrangement of the foliage varies independently at the 
same time ; so that where the floral symmetry of a plant 
does not tally with the leaf arrangement, the discrepancy 
may be due either to subsequent changes occurring in the 
flowers or in the leaves, or perhaps in both. 

For example, a quaternary floral type may be, and often 
is, associated with alternate leaves ; where there is reason 
to suspect that the former was established from a primitive 
opposition in the leaf organs, but that the foliage has subse- 
quently differentiated into a spiral arrangement, leaving the 
original 4-merous symmetry of the flowers unaffected, as in 
many of the Onagracece ; Epilohium, indeed, often furnishing 
ocular demonstration, as, while the lower leaves may be 
opposite, the upper are often alternate. 

On the other hand a quinary arrangement is often 

associated with what may be called a persistent opposition 


in the leaves, as in Carijophylleca and Labiatce. This may 
be due either to an abrupt change from opposite leaves or 
bracts to a spiral one in the flower, or by a reversion 
from an alternate to an opposite position of the leaves, the 
floral organs retaining the arrangements due to their spiral 

The symmetry is based on Calyx, Corolla, and in many 
cases the Androecium also ; but the carpels are not generally 
regarded, for it does not usually extend to the gyncBcium, 
though it is very frequently retained in the andrcecium, 
which is often some multiple of that of the perianth 

In presenting the reader with what may be regarded as 
ostensible grounds for the interpretation proposed, attention 
will be first directed to the more obvious correlations be- 
tween floral symmetry and leaf arrangements, as appear 
from certain numerical proportions ; and, in the next chapter, 
to significant facts observable in the symmetry of particular 

- Commencing with genera possessing alternate leaves and 
a quinary floral type, the prominent fact becomes at once 
apparent that this correlation far exceeds in numerical 
proportion any other. Thus, of above eighty Dicotyledo- 
nous orders * examined in all, no less than 1285 genera 
have quinary flowers associated with alternate leaves, and 
this is exactly what one would expect according to the 
theory advanced that 5-merous whorls are cycles of the 

f type. 

As a corroboration is the fact that such whorls often 
have their parts arranged quincuncially in aestivation (Fig. 
3, a) ; and when they are not so they can be referred to 

* I consulted the first volume of the Genera Plantarum for this 
purpose, which embraces the Thalamifiora: and Calyciflorce. 


it, as I have explained elsewhere : * thus Fig. 3 shows how 
the varieties of imbricate aestivations are deducible from 
the f type (a), bj shifting the edge of the 2nd member under 
the 4th (b, " vexillary "), the 3rd under the 5th (c, " imbricate 
proper"), and the 1st under the 3rd (d, "contorted "). 

Similarly ternary or trimerous whorls are almost universal 
amongst flowers of Monocotyledons, and the ^ type of phyllo- 
taxis is equally common in the foliage. It has been seen 
that the ^ type cannot be deduced from opposite leaves, and 
consequently never occurs, as fur as I know, amongst the 
foliage of Dicotyledons. The comparatively few genera in 
this class with ternary flowers is therefore in accordance 
with the views herein expressed; and where they occur, as 


a 6 o d 

Fig. 3.— Floral ^stlvaticns. 

in Berheris, there are special features which lead one to 
believe they are not due to the ^ type at all, but to the 
breaking up of a high continuous spiral into groups of 
threes, as will be explained hereafter. 

If, however, we take a theoretical departure from a 
single cotyledon, as occurs in Monocotyledons, then the next 
leaf can be at either of the limiting positions of the angular 
distances of 180" or 120'', but not less ; for if it were less 
than 120°, there would be four leaves in any projected 
circle, and this would immediately introduce a member of 
the series ^, ^, f , etc., as shown above. The consequence is 

* See my paper, On the Origin of Floral Estivations, Trans. Lin. Soc, 
2nd series, Botany, vol. i. p. 177. 


that the J and |- types are exceedingly common in the foliage 
of Monocotyledons, while the -|, as far as I am aware, is 
entirely wanting in that class, whether in foliage or flowers. 

Of genera having alternate leaves bat associated with a 
binary or quaternary floral symmetry, there are about 270 in 
aumber of about 30 orders. Now, the co-existence of alternate 
leaves with 2- or 4-meroiis flowers appears at first sight to 
negative the theory ; but, as mentioned above, these and 
other irregularities have been brought about by subsequent 
differentiations in the foliage or flowers. On the other hand, 
opposite leaves with quaternary flowers are not at all in- 
frequent, though not quite so common as when they are 
alternate ; thus, Oleacece and Onagracem are so conditioned. 
Again, in liosacece, which is an order characterized by having 
alternate leaves and 5-merous flowers, three genera alone 
out of seventy have opposite leaves, and these three also are 
accompanied by 4-merous flowers ; viz. BhodotypuSy Coleogyne, 
and Eucrypliia. These three genera thus acquire their 
importance from being isolated amongst others to which 
they are allied, and which are generally otherwise charac- 
terized. Many orders have both foliage and floral symmetry 
remarkably inconstant, and all four combinations, viz. 4- 
merous and 5-merous flowers with opposite or alternate 
leaves almost indiscriminately, as in the tribes Biosmece and 
BorosmecB of Butacece ; and it is a noticeable fact that, 
associated with this inconstancy of correlation, there is an 
inconstancy in the leaf arrangement, opposite and alternate 
leaves being often in species of the same genus, and even on 
the same individual plant. 

The total number of genera noticed as having 4-merous 
flowers and opposite leaves was 110 in 25 orders ; whereas 1 
noticed 276 genera of 30 orders as having 4-merous flowers 
associated with alternate leaves. This, I believe, is due to 


subsequent differentiation in tlie foliage to an alternate 
condition, the quaternary condition of the flowers remaining 

Similarly with the last condition, I found 212 genera of 
30 orders with a quinary arrangement of the flosvers corre- 
lated to an opposite condition of the leaves, this being an 
apparent anomaly of the same kind, but which is, however, to 
be interpreted in the same way. Thus the Lahiatce are con- 
stantly 5-merous in the flowers, but with as constantly 
opposite leaves. Now, if we contrast this order with Scro- 
'phidarinecB, we find a similar constancy in certain genera 
only, as in Bliinantlius^ etc. ; while other genera have alter- 
nate leaves as Linaria, Digitalis, etc. 

There is an alternative of interpretations of this fact, for 
both can be illustrated in nature. Either all the pentamerous 
flowers have been deduced from alternate leaves (as may 
have been the case with lihinantJius and Lahiatce), the leaves 
having subsequently reverted to the original or ancestral 
state of opposition; or else, the 5-merous character of the 
flowers has arisen by a sudden change (possibly due to the 
stimulus of insect agency) from opposition in the leaves or 
bracts to an alternate arrangement in the parts of the flower. 
As an illustration of this latter process may be mentioned 
the development of the five sepals of Beutzia as compared 
with the four of the allied genus PhiladelpJuis. In this 
latter genus the anterior and posterior sepals appear 
together, subsequently the two lateral arise simultaneously. 
In Deutzia, however, the two anterior sepals correspond to 
Nos. 1 and 3; two sepals are lateral, viz., Nos- 4 and 5; 
and the posterior sepal is No. 2. Thus the opposite and 
decussate pairs of sepals of PhiladelpJuis w^ould be repre- 
sented by the figures 1 and 2, 3 and 4. If these were to 
break up into a quincuncial spiral and shift their positions. 


they would, with the addition of one more sepal, assume 
those represented by Deutzia. 

2— > 

Exactly the same procedure occurs in the change from 
opposite to alternate arrangements of leaves in the Jeru- 
salem Artichoke, as I have explained in treating of the 
varieties of leaf- arrangement in that plant. 

Calycanthus is another instance illustrating an abrupt 
change from an opposite condition of the leaves to the ^V 
type in the bracts enveloping the flowers, and which then 
pass insensibly into sepals and petals. 

Symmetrical Increase and Decrease in Floral Whorls. — 
As another instance of variability adding further complica- 
tions, it may be observed that in both kinds of arrangements, 
namely, of those plants possessing alternate and those pos- 
sessing oi3posite leaves, there are many genera whose floral 
symmetry ranges from one to some higher number in the 
different species of the same genus. Thus 4-5-merous flowers 
are especially common. I found it so in more than 100 genera 
of 23 orders examined among alternate- leaved plants ; and 
58 genera of 19 orders among those with opposite leaves. 

Again, some genera have species the whorls of whose 
flowers range from 3 to 5 or 6, or from 4 to 6 in the number 
of parts ; others from 5 to 7 or 5 to 8, etc. In these 
cases it is often quite impossible to explain what has been 
the immediate causes producing such variations. The only 
interpretation that can be given is that the primary sym- 
metry having been originally determined by phyllotaxis, it 


changes, whether in the individual or in its descendants, 
through the law of " symmetrical increase or decrease." By 
this I mean that the number of sepals, petals, and stamens 
often vary together from the typical number by the addition 
or subtraction of a member. Thus, in a single corymb of 
an Elder, 4-, 5-, 6-merous flowers may be often found; simi- 
larly, while early blossoming Fuchsias may bear 3-merous 
flowers, they are replaced later by the regularly 4-merous 
ones. Although these changes frequently occur in the same 
plant, they usually are not permanent. Yet they occasionally 
appear to have become so, as in the terminal flowers of Adoxa 
and Monotropa. On the other hand, the constant occurrence 
and, therefore, specific character of 4-merous flowers in 
Potent ilia Tormentilla, and 3-merous in Tillcea onuscosa, 1 
should be inclined to attribute to the fixation of a symme- 
trical reduction which has taken place from the permanent 
5-merous type so characteristic of Potentilla, and many 
genera of the Crassidacece. Not infrequently the difference 
of number is pronounced by systematists as generic; thus, 
while Rahia has 5-merous flowers, Galium has 4-merous. A 
similar difference lies between Buta and Haploj^hyllum* 

If a cause be looked for, it would seem to be merely a 
question of nutrition. If the symmetry varies in the same 
plant, it is obvious that a corolla of four petals could not 
have been provided with the same amount of nutritive 
material as a 5-merous one. But if it be a specific character, 
as in Tormentil (which, it may be observed, affects the more 
or less barren soil of heaths), then the change has become 
fixed and is now hereditary. 

* Bt running the eye through the artificial keys at the commence- 
ment of the Orders in the Genera Plantarum of Bentham and Hooker, it 
will be seen how frequently these authors regard the number of parts in 
the Calyx and Corolla as a prominent generic character. 


Unstmmeteical Decrease in certain Floral Whorls. — 
Another modifying cause of the change of symmetry id 
the adaptation to insect or other agency for fertilisation. 
This I believe to have played a most important part in modi- 
fying flowers, as will be explained more fully hereafter, more 
especially in affecting the Androecium and Gynoecium, than 
the Perianth, as far as "number" is concerned, this latter 
organ being altered by their agency, more especially in Form. 
Thus, the loss of one ur more stamens is very characteristic of 
certain groups, as in the Labiatce, when the remaining mem- 
bers of the androecium become altered in length and position 
so as to facihtate the intercrossing of distinct flowers. 

On the other hand, with inconspicuous and cleistogamous 
flowers, there is a strong tendency to reduce the number of 
stamens, as in Cbickweed to three, the allied species Stellaria 
Holostea having ten. Similarly, in the cleistogamous flowers 
of Violets they are sometimes reduced to three or two ; since 
a very small amount of pollen is really quite sufficient to 
fertilise a considerable number of ovules. 

The gynoecium has very frequently a less number of 
carpels than the other whorls have parts. Now, the primary 
effect of intercrossing is to enhance the size of the corolla 
and to give a preponderance to the androecium. On the 
other hand, one result is to check for a time the growth and 
development of the gynoecium of most insect-visited herma- 
phrodite flowers, i.e. to render the flower protandrous ; and 
I strongly suspect that the generally reduced number of 
carpels in highly differentiated flowers — as of the Gamojpetalce^ 
in comparison with the Thalamiflorce and Calycifioroe — is cor- 
related to the fact that they have been for many generations 
visited by insects. This idea is supported by the fact that 
bicarpellary genera sometimes tend to restore the ancestiul 
number of the five carpels, as is occasionally the case in 
Gesneria . 


In some cases, nature seems, as it were, to try and com- 
pensate for the loss of the carpels by an increase in the 
quantity of seeds. Thus, while no Labiate flower has more 
than four seeds, it has been ascertained that a Maxillaria 
bore 1,700,000 seeds ; and I found by calculation that a single 
plant of Foxglove yielded a million and a half apparently 
good seeds. 

The relative advantages of having many or few seeds 
will be discussed later on. 

Illustrations from Raxunculace^. — Certain genera of 
the Eanimcidacece are particularly instructive in showicg 
how members of the floral whorls originate in phyllotactical 
methods, but are more or less 
altered in their positions by the 
lateral union of their fibre- vascular 
cords ; so that they become ar- 
ranged in superposition instead 
of being alternate, or vice versa. 
Thus, in Garidella (Fig. 4) (with 
which Hellehorus fcetidus partly 
agrees), the sepals and petals are 
both arranged, and arise succes- 
sively, in quincuncial order; the ^ 
petals being (correctly, in accord- '^' '" '^sram o 
ance with phyllotaxis) superposed to the sepals. The an- 
drcBcium forms a Avhorl of eight stamens, and represents 
a cycle of the f arrangement ; the proper angular divergence 
of 135° is, however, not retained, in consequence of the fibro- 
vascular cords being intimately connected with those of the 
petals. Having thus established the first whorl of eight, the 
rest of the staminal series follow on the same radial lines. 
By referring to the diagram (Fig. 4) it will be seen how the 
stamens of the outermost whorl group themselves in super- 



position to the petals and sepals. Similarly, in Nigella 
sativa tlie petals are eight in number, and occupy the same 
positions as the outermost whorl of stamens of Garidella. 
They have, then, the eight stamens of the outermost whorl 
of the androecium superposed to them. 

In Delphinium the stamens and carpels form a continuous 
spiral, represented by |, or approximately by f. In some 
cases Braun* found 16 stamens, and the first carpel being 
the 17th organ, stood superposed to the stamens No. 9 and 

No. 1. In another case 18 
stamens were developed, so 
that the first carpel stood 
superposed to stamen No. 11. 
Helleborus niger (Fig. 5) 
has five sepals which emerge 
and are arranged in quin- 
cuncial order. There are 
twenty-one nectariform pe- 
tals, i.e. one cycle of the 
■^-j- arrangement, grouped as 
in the accompanying dia- 
gram. The petals 1 to 8 
and 9 to 16 would correspond approximately to two cycles 
of the f type. Radial rows of stamens then follow on the 
same lines as the petals. 

Eranthis hyemalis has, as usually regarded, a 5-8-merous 
coloured calyx. A pair of staminodes stand superposed to 
each member of the outer whorl. Stamens follow along the 
radial lines, of which six terminate in carpels. 

Aguilegia vulgaris, or the Columbine, has the sepals, as 

* Al. Brann on Delphinium (Pringsheim's Jahrl. f. Wiss, Bot., 1857, 
i. 206), referred to by Henfrey, Morphol. of BalsaminecB, Journ. of Lin. 
Soc, iii. 159. 

Fig. 5.- Diagram of Helleborus niger. 


usual, quincuiicially arranged. The petals appear simul- 
taneously^ alternating in position to tlie sepals. The stamens 
occur in ten rows, 5 being superposed to the petals and 
5 to the sepals ; and, lastly, 5 carpels appear superposed 
to the petals. This flower, then, adopts the more usual 
character of alternation in the whorls. But it may be noticed 
that while the corolla alternates with the calyx, each of 
these outer whorls gives rise to a radial series of stamens. 

From the preceding illustrations, it will now be seen that 
phyllotaxis lies at the foundation of the arrangements of the 
members of floral whorls ; that the -| type prevails in the 
sepals and petals, with a strict angular divergence of 144°. 
The divergences are, however, subsequently modified in the 
stamens and carpels. Thus, in Hellehorus niger the petals 
clearly represent a whorl of 21 parts, i.e. they are pre- 
sumably arranged according to the 2t ^JP^- They are, how- 
ever, so far modified in position as to become superposed 
to the sepals in groups. Similarly the stamens form 
series of 21, each being superposed in radial lines to the 

The interpretation of these displacements from what 
would be due to strict, phyllotactical laws is that the 
individual cords of the stamens and carpels are not inde- 
pendent as they are in the " leaf traces " of an axial cylinder, 
where the cord or cords belonging to each leaf are simply 
intercalated side by side with those of the leaves most nearly 
approaching the same vertical line, and constitute together 
the common fibro-vascular cylinder of the stem. In the 
pedicel, however, the rule is that this should contain at least 
the same number of cords as there are leaves to the perianth, 
or sepals and petals together. These, usually six or ten 
cords, on reaching the floral receptacle are sent off respec- 
tively as the cords of the sepals and petals ; whereas, it is 


tliese latter v^h.\c\i by lateral or radial " chorisis " supply the 
cords required for the stamens and carpels 

The consequence is that the essential organs have their 
cords issuing from a common stem Avith those of the perianth. 
Thus they are compelled to stand superposed to them. 

Perhaps the word "compelled" requires a word of ex- 
planation. The cord of any organ superposed to another 
may be given off either by radial, i.e. lateral, or tangential 
chorisis from the cord of the latter. Instead, however, of 
the new lateral branch giving rise to an organ by the side of 
the former, it results, partly from the close proximity of the 
two and partly from the tendency of the remaining cords of 
the cylinder to " close up," that the new member finally 
takes up a position in front of, i.e. superposed to, the one 
whose cord has given rise to it. When a cord is separated 
by tangential chorisis, as is so often the case with staminal 
cords, then the resulting organ must necessarily be super- 
posed to the one, from the cord of which it has been 



Illustrations of Special Numbers. — It will now be advisable 
to give examples of particular numbers occurring in flowers, 
and attempt to account for them. 

One-membered Whorls. — Where one part to a whorl is 
only found, it may in nearly every case be regarded as a 
degradation from some higher number. The only instances 
I am aware of in which the calyx seems to consist of a 
single member are some species of Aristolochia. In Musscenda 
one out of the five sepals is greatly enlarged to become an 
attractive organ.* 

One petal is occasionally found. Thus, four genera of 
Vocliysiacece have each only one petal to their flowers ; but 
as the sepals are five in each of the seven genera of this 
order, and the petals range from one to five in number, the 
inference is clear that the solitary petal of these four genera 
is due to the arrest of the others. 

One stamen occurs more frequently ; as in Hippuris, 
Gentrantlius, JEuphoi'hia, Casuarina, OrcMs, Canna, Lilcea, 
Lemna, etc. As allied genera have more than one, and it is 
accompanied by other signs of degradation or metamorphosis, 

* If there be one external foliar organ only, it is regarded as a bract, 
as in Willows and Aponogeton. 


there is no doubt but that similar processes will account for 
one stamen as for one petal. Thus Sippuris with one, is 
allied to Myriojpliyllum with four ; while Centranthus has one, 
Fedia has two and Valeriana three. 

Casuarina alone seems to raise a doubt of its being 
degraded and possibly a primitive form ; but this is solely 
because it has no living allies (excepting perhaps Myrica). 
The terminal stamen would not be of itself a point of 
importance, as it has a parallel in Euphorbia ; but it is its 
isolation without affinities, its peculiar equisetum-lrke habit, 
which seem to indicate great antiquity, so that no inference 
can fairly be drawn to interpret its present monandrous 

Amongst Monocotyledons, Canna is clearly monandrous 
by petalody of the other stamens. Orchis by metamorphosis 
also. Lastly, Naias, Gaulinia^ Zostera, Zannichiella, and Lemna 
are in all probability greatly degraded forms from higher 
plants, degradations being the usual effect of an aquatic life, 
and not primitive types of Monocotyledons. 

One carpel is not at all uncommon, as in the Leguminosce. 
As Affonsea has five, the absence of four in this order is no 
doubt due to arrest. In the tribe Berherece, however (if my 
interpretation be correct, of the origin of the seven whorls 
of three each constituting the flowers of Berheris, as explained 
below), the one carpel may be the last of an originally 
continuous spiral, formed from eleven pairs of opposite leaves, 
now broken up into seven ternary whorls, with one over. It 
may, however, be the remaining one of three, which possibly 
constitutes a ternary gynoecial whorl, which is characteristic 
of the tribe Lardizahaleoi of the same order Berheridece. 

Dimerous Whorls. — A dimerous arrangement is not par- 
ticularly common, though a quaternary calyx is dimerous in 
its development, as the sepals emerge from the axis in sue- 


cessive pairs.* The following may be taken as illustrative 
instances. The sepals of Fapaver and Fumaria, the outer 
stamens of Cruciferce. In Circcea all the whorls are dimerous, 
in Oleacece the essential organs alone, as also in Pinguicula, 
Salvia, Veronica, and Salix diandra. 

The question arises, is this number two an original one, 
or has it arisen by arresting some parts of a more numerous 
whorl ? It is obviously so with Salvia and other genera of 
the Lahiatce, where rudimentary stamens are present. So 
also with Senehiera didyma where the two stamens take the 
place of the four larger ones of other genera of the Cruciferce. 
It is probably so with the two imbricate sepals of Poppies, 
those of P. orientale being often increased to three, which 
seems to be a tendency to revert to a more primitive and 
higher number. 

With such plants, however, as Circcea,. the Ash, and Vero- 
nica^ which have retained opposite leaves, the dimerous 
whorls may be a primitive condition. This idea is ostensibly 
supported by the fact that the outer whorls of the flow^ers 
are quaternary and not quinary, since, when this is the 
case, the sepals always issue in pairs from the axis, and not 
simultaneously as do the petals ; but as long as no rudi- 
mentary organs exist, there is nothing to disprove the idea 
that in these genera the number of stamens may not be due 
to degradation. Indeed, all analogy would lead one to 
suppose so in most cases, as of CirccBa and Veronica: the 
binary whorls of the former genus, and the quaternary outer 
and binary inner -whorls of the latter, being presumably due 
to " symmetrical reduction " from the prevailing quaternary 

* Though the antero-posterior sepals of crnciferous flowers are 
regarded as the most external, it is really the lateral ones which are 
first provided with fibro-vascular cords from the complete oblong 
cylinder in the pedicel, just as in Cleome (see Fig. 6, p. 32). 


type of the Onagracece and quinary of the Scrophularinece 

Trimerous Whorls. — The number three is strongly 
characteristic of Monocotyledons, and appears to be in this 
class the immediate result of the J phyllotaxis. In Dicoty- 
ledons, however, there are certain orders in which it prevails, 
and it will be noticed that the number of parts in those 
orders is generally much increased ; as in Magnoliacece* 
AnonacecB, Berheris, Laurics Campliora, Rumex, etc. In some 
the andrcecium and gynoecium are so increased in number 
that they cease to be whorled, but have become spirally 
arranged on a more or less elongated receptacle and are 
represented by the fractions y^ or 2t-* 

It has been demonstrated above that a pentamerous 
arrangement is undoubtedly due to the | phyllotaxis, each 
whorl constituting a cycle ; but if the fraction be a higher 
one, as y% or ^'^y, then the number of parts in a cycle are too 
great to be compressed into a whorl. Nature appears then 
to adopt another method. Falling back upon the law that 
with these arrangements no part of the continuous spiral, of 
sufficient length to constitute a complete circle when pro- 
jected upon a plane, ever contains more or less than three 
leaves (excepting the ^ type), the series is now broken up 
into a succession of ternary whorls, the whole forming the 
complete flower, and, being taken together, corresponds to 
about or exactly one cycle of a high type. Thus Barberry 
has 3 bracts, 3 -f 3 sepals, 3 + 3 petals, 3 -j- ^ stamens 
and one carpel ; that is, seven whorls of threes or twenty-one 

* In Magnolia an individual complication is introduced, in that the 
immense number of stamens and carpels is secured by doubling the 
whole number attributable to the j% arrangement. Consequently, instead 
of there being five and eight " secondary spirals," there are ten in one 
direction and sixteen in the other. 



'parts, and one over. If these seven whorls wel 
and arranged spirally, they would be represented 
then there would be eight coils in the cycle. The 
of seven and not eight whorls is due to the fact that in 
rearranging them, so to say, in a verticillate manner, and by 
necessarily shifting the position of the parts, a certain por- 
tion of the spiral line is lost in forming each whorl, as the 
angular divergence between two parts in a whorl is 120°, 
but on the spiral it is nearly 123° ; so that by the time the 
twenty-first organ is arrived at, only seven circles have been 

Similarly, in Bumex, if we supply the theoretically lost 
corolla, the flower would consist of twenty-one parts exactly.* 

Another and somewhat frequent origin of the number 
three in Dicotyledons is due to what I have called sym- 
metrical reduction : when not only the different species of a 
genus may have the number of parts of their floral whorls 
ranging from 5 to 4 or 3 ; but such variations may occur on 
the same plant. Thus Butacece (following the Gen. Plant.) 
has 34 genera with 5-merous flowers ; 18 genera with species 
varying from 5 to 4-merous ; 16 are 4-merous ; 3 range from 
5 to 3-merous ; 2 from 4 to 3-merous, and 1 is 3-merous. 

Tetramerous Whorls. — That a true quaternary arrange- 
ment is due to an opposite condition of the foliage seems 
borne out by statistics, though quinary flowers are not at all 
uncommon as well. Thus of Butacece there are 6 genera 
with opposite leaves and 4-merous flowers; 2 only with 
5-merous, and 2 with 4-5-merous flowers. On the other 
hand, there are 25 genera with alternate leaves and 5-merous 

* High spirals can be otherwise treated, as in the case of Chimonan- 
thus, where whorls of fives are made out of a spiral system of ^j (see 
below, p. 38). 



Another correlation witli a quaternary arrangement is a 
not nnfrequent valvate condition of the sepals at least, or of 
the sepals and petals as well. These conditions prevail, for 
example, in Oleacece, Onagracece, and, with the exceptional 
genus, Clematis, of the Banunculacece. Too much stress 
must not be placed upon this coincidence, as, if the petals 
be enlarged through insect or other agency, the valvate 
aestivation is often lost, and the petals become imbricate, as 
in Fuchsia, Godetia, etc., though it is there retained in the 
sepals. This valvate condition is foreshadowed in the ver- 
nation of the foliage ; in that opposite leaves are almost 
invariably valvate, having the two upper surfaces of the 
leaves pressed together, as may be seen in Hypericum and 
Vinca ; or else with the edges induplicate, as is characteristic 
of Gaprifoliacece, resembling the sepals of Clematis.* 

Though the Onagracece have a preponderance of genera 
with 4-merous flowers, there is in this order great variation 
in the foliage. It is strictly opposite in Fuchsia and others, 
but 14 genera out of a total of 22 have alternate leaves, while 
with some, like Epilohium, it varies on the same stem. This, 
I think, reveals the fact that the 4-merous condition has been 
first established in the flowers, and subsequently the foliage 
has varied from an opposite to an alternate condition in 
certain genera, just as it does in an individual plant of 

That symmetrical reduction has elsewhere played an 
important part in the origin of 4-merous flowers, is a sup- 
position fully borne out by facts. In some cases it has 
seemingly established itself as a permanent character, so that 
systematists recognize it as generic or specific, accordingly, 

* See a paper by the author, On Vernation and the Methods of 
Development of Foliage as protective against Radiation, Journ. Lin. Soc. 
Bot., voL xxi., p. 624. 


as the case may be ; this, Haplophyllum may be compared with 
Buta, Buhia with Galium, or, again Potentilla reptans with 
P. Tormentilla, etc. On the other hand, I repeat, when one 
observes that of the /'I genera of Rosacece three only are 
recorded in the Gen. Plant, as having opposite leaves, and 
these three are characterized as having 4-meroas flowers, viz. 
Bhodotypus, Eucryphia, and Coleogyne, there appears to be 
a significant correlation between quaternary flowers and 
opposite leaves. 

A quaternary arrangement is found very exceptionally 
in Monocotyledons, as in the order Naiadacece, e.g. Tetron- 
cium and Potamogeton. As the numbers 6 {i.e. 2 X 3), 4, 
2, and 1 are found in different genera, the quaternary 
as also binary arrangements may, I think, be reasonably 
referred to symmetrical reduction. 

Perhaps of all orders the quaternary arrangement (at 
least in part) of Crucifers has raised more discussion than 
any other kind of floral symmetry.* 

Without entering here upon any lengthened discussion 
I would only add that, as far as investigations into the 
anatomical structure of the pedicel is concerned, there is a 
decided difference from what occurs in most flowers having 
a definite number of parts, and where the whorls are 
regularly superposed to one another, in that the members 
of the whorls not being for the most part on common radial 
planes, they have not their cords fused to^*ether in the usual 
manner in a radial direction. 

A section at some distance below the flower reveals four 
or five cords forming a circle. These rapidly increase in 
number by branching laterally, till between ten and twenty 
are found arranged in an oval just below the flower. Two 

* See my paper On the Structure of a Cruciferous Flower, Trans. 
Lin. Soc, 2nd series, Botany, vol. i. p. 191. 



cords, one at each end of tlie long axis, now part company 
from the rest, and enter the lateral sepals (Fig. 6 (a) Z.5.), 
the antero-posterior sepals next receiving their cords (a.s. 
andjp.s.). The cylinder tends to close np, and four groups 
situate at the corners of the oblong cylinder supply cords for 
the petals, p. The two honey-glands next put in an appear- 
ance, G. They are merely cellular expansions of the floral 
receptacle, and are entirely devoid of cords, and therefore 
not rudiments of appendages. The two lateral stamens next 
receive their cords,, while four other cords are given off 
from beside the petaline for the taller pairs of stamens, st. 

^ m.c. 

Fig. 6.— Anatomy of Wallflower. 

Fig. 6 (6) shows how their cords diverge below and spring 
from the siae of the petaline cords, while extra cords arise 
between them to form the marginal cords of the carpels (m.c). 
From this it will be seen that the longer stamens cannot be 
formed by " chorisis " of a common intermediate cord ; but, 
like those of all other members of the flower, their cords are 
sepai-ated from the common fibro-vascular cylinder of the 

The conclusion suggested by this investigation, and by a 
comparative study of Capparidece, is that a cruciferous flower 
is not reducible to an originally quaternary type at all, but 
to some higher one. In my paper referred to, I suggested a 


quinary ; but I am now more inclined to refer it primarily to 
an indefinite spiral series referable to the ^^ or ^ type, 
which has been reduced, perhaps through insect agency, by 
symmetrical reduction to the present anomalous condition. 

The process of transition from a hypothetical indefinite 
number of stamens to the present hexandrous state may be, 
perhaps, seen by comparing the three genera of Capparidece 
— Cajpjparis, Polanisia and Cleome. The first has many 
stamens and six placentas, which are sometimes reduced to 
two. Polanisia has eight stamens, or more rarely six. Their 
situations correspond exactly with those of the Cruciferce, 
except that, when there are eight, there are four on the 
anterior side instead of two. 

Lastly, Cleome brings us to the same structure as in the 
Cruciferce with even the tetradynamous condition of the 
stamens ; the elongated torus below the pistil being about 
the only " capparidaceous " feature left. 

It is not at all uncommon to find more than six stamens 
in cultivated plants of the Cruciferce, and when this is the 
case I should be inclined to regard it as a tendency to a 
reversion to a higher ancestral number. 

On the other hand, the close proximity of the two taller 
ones on each side not infrequently brings about some degree 
of cohesion between them, with an occasional arrest of half 
an anther. This has led some to suppose that the pair have 
resulted from chorisis. Since, however, their cords diverge 
downwards to the right and left, and run down beside the 
petalline cords (Fig. 6, 5), this clearly proves that the union 
is a result of close contact, and that the normal separation is 
not due to chorisis, but to a primitive freedom, which has 
been retained from a multistaminate condition. 

Pentamerous Whorls. — These are by far the commonest 
amongst Dicotyledons. And as an enormously greater pro- 


portion of plants in this class have alternate leaves and 
5-merous flowers, this correlation alone would be almost 
sufficient to prove that the latter issued out of the com- 
monest or f type of phyllotaxis. But since the sepals are 
sometimes decidedly quincuncial, as are those of Digitalis, 
and the petals frequently so, we have undoubted proof that 
they represent cycles of this angular divergence. 

As with other numbers, fives may arise by symmetrical 
increase from fours, or decrease from sixes ; though in by 
far the greater number of instances it is a primitive number, 
as stated above. As a rare instance of symmetrical decrease 
may be mentioned Lythrum Salicaria, which has usually the 
central floret of each axillary cyme 6-merous, but the lateral 
ones only 5-merous. As an instance of five parts to a whorl 
amongst Monocotyledons, may be mentioned the stamens of 
Strelitzia regina ; but this number is obviously due to the 
suppression of a stamen. 

Although whorls of fives are cycles of the f divergence, 
and usually follow after an alternate arrangement in the 
foliage, yet it is quite possible to change abruptly from 
opposite leaves or bracts to whorls of fives in the flower, as 
may be seen in Hypericum and DiantJius. This arrangement, 
as I have elsewhere shown, is that most easily acquired 
when opposite and decussate leaves become alternate by the 
development of internodes (see pp. 11 and 18). 

Hexamerous Whorls. — A floral whorl of six parts is, in 
most cases, as amongst Monocotyledons, the result of the 
combination of two whorls of three each — as the androecium 
of Berheris, Tulip, or perianth of the Lily of the Valley. It 
may, however, arise from symmetrical increase, as, for 
example, in the orders Meliacece and Olacineoe. In the 
former, there are 18 genera with alternate leaves and 
5-merous flowers ; 9 with 4-5-merous ; 4 with 4i-merous ; 


4 with 5-6-merous, and 1 witli 4-6-merous whorls in the 
different species. In Olacinece, of 36 genera, 17 have alter- 
nate leaves and 5-merous flowers ; 7 have 4-5-merous ; 4, 
6-6-merous ; 2, 6-merous, and 1, 4-6-merous. 

As six leaves cannot form a cycle of any of the ordinary 
kinds of phyllotaxis, this will account for its rarity in 
nature ; and indeed it may probably, without exception, be 
divisible into two whorls of three members each, except in 
the case of symmetrical increase from five. 

Heptamerous Whorls. — Like the number 6, 7 is a very 
rare one ; and when present appears to be due to its being a 
primitive number or to symmetrical change. If any whorls 
are deducible from decussating verticils of threes, a cycle 
may contain seven parts, as the phyllo tactical series arising 
from the breaking up of such verticils into a continuous 
spiral arrangement is represented by -J, ^, -f-, y^, etc. So that 
if leaves on a plant were in whorls of threes, as occurs in 
some instances, and not opposite, as in the primitive type 
amongst Dicotyledons, then a heptamerous arrangement 
would occur. If, therefore, there be any existing illustra- 
tion, it must, by the very nature of the case, be exceedingly 
rare. It sometimes occurs in Trientalis ; and when this is 
the case, it may possibly have arisen as here suggested. 
According to the description given of this plant in the 
Genera Flantarum, the numbers of the three outer whorls 
range from 5 to 9, the capsule being 5-valved. The leaves, 
on the other hand, are " saepe tot quot petala subverti- 

A second cause is arrest. This obviously accounts for 
the 7 anthers in Pelargonium, for the 10 filaments are present. 

A third cause is symmetrical change. Ly thrum Salicaria 
illustrates this as already mentioned. This flower is some- 
times described as 6-merous, but it is not always so. The 


central floret of the cyme has often a higher number than that 
of the lateral ones ; so that if they be 6-merous, the central 
flower will be 7-merous. Agapanthus, amongst Monocoty- 
ledons, is another instance, its flowers ranging from 6 to 8 in 
the number of parts in the whorls. 

OcTAMEROUS Whorls. — A whorl of eight parts is not 
common ; but it appears in Chlora and in the corolla of Dryas 
octopetala, in which it may be a cycle of the f phyllo taxis. 
In other cases it is a combination of two whorls, which, as a 
rule, can be easily distinguished as the stamens in the Ona- 
gracece, or it may be due to symmetrical change. 

Enneamerous Whorls. — The number 9, like 6, 7, and 11, 
corresponds to no cycle of any one of the usual forms of leaf- 
arrangement, and is proportionately rare. It may occur as a 
combination of three cycles of three each, and perhaps this 
will account for it when it occurs in Trientalis, and the 
androecium of Mercurialis. The stamens of Butomus are 
also nine in number. 

Decamerous Whorls. — The number 10 never occurs 
except as the union of two whorls of five in each, as in the 
androecium of Leguminosce. 

Endecamerous Whorls. — Like 7, the number 11 might 
occur if the series ^, i, f, -j^, etc., was as frequently repre- 
sented as ^, ^, f , f , etc., when " sevens " would be as abun- 
dant as " fives " are now. I do not know of a case where it 
could reasonably be referred to such an origin. When it 
does occur, as in Cuphea, it is clearly due to an arrest of one 
stamen through insect agency. Broivnea is said also to have 
sometimes 11 stamens ; if so, this would undoubtedly be due 
to numerical increase. 

DoDECAMEROUS Whorls. — The number 12 closely verges 
on the " indefinite," which simply means a more or less 
numerous series of cycles of the same kind. Neverthe- 


less, it occurs as a " definite " number in several instances. 
The 12 stamens of Lythrum are, of course, two series of six 
each. Both 12 and 24 are found in the Crassulacece, as in 
Sempervivicm, in which genus the petals vary from 6 to 20, 
and the stamens from 12 to 40. This seems to show that in 
the one case they are combinations of cycles of threes, in the 
other, of fives ; just as Berheris illustrates the former, Cliimo- 
nantlius the latter instance. 

Indefinite Whorls. — As soon as we pass from twelve to 
some higher number, then flowers cease to be whorled, and 
the parts are arranged spirally, and follow more or less 
exactly the laws of alternate phyllotaxis ; interferences occur 
in consequence of the want of space, some secondary spirals 
being often incomplete. Moreover, since the fibrovascular 
cords become fused, in other words branch by chorisis, and 
are not independent as of ordinary foliage, parts take up 
slightly different positions to what they would if they could 
strictly follow phyllotactical laws. 

I have alluded to what I call " symmetrical increase and 
decrease " as causes of variation in the number of parts of 
whorls ; and what brings about these variations in number, 
is an excess or deficiency of nutriment and vital activity 
respectively. There are innumerable examples of all the 
above kinds of changes in number. In fact, if any one or 
series of whorls of a flower be w-merous, it may become 
n ± oj-merous, and will give rise to symmetrical increase or 
decrease accordingly ; or again, three whorls of the same 
flower may become n ± x, n ± y, n ± ^-merous ; when all 
numerical symmetry between them will be destroyed. 

Similarly, if the parts be spirally arranged, the number 
may vary from the prevailing one by increasing or decreasing 
the length of the spiral, both in flowers of the same plant or 
in different species of the same genus ; as, for example, may 


be seen by comparing the number of stamens in a large- 
flowered form of Eanunculus aquatilisj with the small- 
flowered B, hederaceiis ; or one genns with an allied one, as 
Banunculics with Myoswus, in which the stamens are reduced, 
often to one whorl of five only. 

Lastly, just as high spirals can be broken up into ternary 
whorls, so can the arrangement gx be separated into whorls 
of a lower series, as of 13, 8, or 5 parts respectively. Thus, 
of the two genera, which have opposite leaves, comprising 
the order Calycanthacece, Calycanthus illustrates an abrupt 
change from opposite leaves to the ^\ arrangement in the 
bract-like sepals of the flower; but no distinction between 
bracts, sepals, and petals can really be made. Chimonanthus, 
however, would seem to be a more highly differentiated type, 
in that, not only is the calyx distinguishable from the corolla, 
but five exterior stamens constitute a distinct whorl by them- 
selves, and the indefinite barren ones of Calycanthus are here 
reduced to five ; so that, omitting the pistil, the flower con- 
sists of four distinct pentamerous whorls. 



m. 1520. n22M.w msf 

Superposition and Alternation of Whorls. — It has been 
already observed that leaves are arranged on two methods, 
either being on the same plane, i.e. opposite and verticillate ; 
or with only one at a node, i.e. alternate. If the fibro-vascular 
cords passing from the leaves into 
the stem be traced downwards, 

those belonging to the leaves ^^ 

situate in one and the same ver- to 
tical line always have their lower ^— 
extremities inserted laterally and 
not actually confluent in that line, 
as will be seen in Fig. 7, taken 
from Hanstein's researches.* 

This fact is true, not only for 
foliage and bracts, but also to 
some extent for sepals and petals. 
When, however, we trace the 

nricrin nf <?fnmpn«! anri pm-npla wp Fig. 7 —Diagram of the foliar cords in 
origin OI stamens ana caipeiS, we the stem of the Arabis aWida (after 

find that their cords, instead of Hanstein). 

being inserted separately into the fibro-vascular cylinder, 

generally arise by branching, or by the so-called " chorisis " 

* De la Connexion qui existe entre la Disposition des Feuilles et la 
Structure de la Zone Ligneuse des Dicotyledons, Ann. des. Sci. Nat., 4® 
ser., torn. 8. 

2.Z 4i 3 J & 3.8^S. 


of the cords belonging to the sepals or petals, or from both ; 
and similarly the dorsal cords of the carpels branch off from 
the same stem as that of the sepals or petals, very rarely 
from both at once. Simultaneously with the dorsal, two 
marginal cords pass up directly into the placentas, having 
originated in the same way ; and, in so doing, the floral re- 
ceptacle usually becomes extinct, and takes, as a rule, no 
further part in the construction of the central portion of the 

Starting, then, with these two fundamental sources of the 
various arrangements of the parts of flowers, we may first 
observe that of opposition or superposition and alternation, 
the former, if represented by decussate pairs of appendages, 
is the most primitive type. This is seen in many quaternary 
flowers in which the sepals emerge in successively decus- 
sating pairs. Such opposite leaves being foreshadowed in 
the cotyledons of exogens. 

The next, or rather the first stage of differentiation is 
seen in the spiral condition which obtains in many flowers, 
mostly represented by the ^ and | types : thus, e.g., ^ repre- 
sents the arrangement prevailing in petaloid Monocotyledons ; 
and all pentamerous calyces issue in a quincuncial manner. 
In Sahia, the petals follow continuously with the sepals in 
the same spiral line, so that the first petal is superposed to 
the first sepal. These whorls accordingly represent two 
cycles of the f type, as seen above in Garidella (p. 21). 

By far the commoner condition is to break up the spiral 
into cycles, say of five parts each, and then to shift their 
positions, so that they become alternate instead of superposed. 
Now, such a decussate arrangement is usually described as a 
fundamental law, not only governing opposite and verticillate 
leaves, but floral whorls as well ; and particular stress is 
laid upon the usual presence of the petaline whorl of carpels, 


inasmucli as the law o£ alternation is thus carried out 
completely, and which may be represented as follows — the 
hyphens indicating the parts superposed to one another — 
Sepal-stamen ; Petal-carpel. 

From what has been stated above, the true order of 
arrangement and superposition would be — Sepal-stamen- 
carpel ; Petal-stamen-carpel ; and either one of the staminal 
and either one of the carpellary whorls may be suppressed. 
Thus, for example, Oxalis, Zygophyllum, Geranium^ and Ruta 
have Sepal-stamen ; Petal-stamen-carpel : while Limnantlies, 
Coriarea, and Agrostemma have Sepal-stamen-carpel ; Petal- 
stamen. As instances where there is but one whorl of sta- 
mens. Campanula and Sermannla have Sepal-carpel; Petal- 
stamen ; whereas Linum and Diosvia have Sepal-stamen ; 

Of these variations, although Sepal- stamen is commoner 
than Petal-stamen, and Petal-carpel than Sepal-carpel,* 
yet these are, so to say, rather matters of accident than 
otherwise, in that it is probably due to certain exigencies of 
nutrition, and especially insect agencies, that such variations 
of arrangement exist. 

The important fact mentioned above, that floral whorls 
are projected cycles and not primitive whorls, has, as far as 
I know, been entirely overlooked by botanists. Thus, for 
example. Professor Asa Gray remarks on the presence of 
whorls in flowers as follows : " Cycles alternating with each 
other are simply that of verticillate phyllotaxy," f to which 
he refers the opposite, ternate, quaternate and quinate verti- 
cils. J In the case of leaves, verticils represent usually more 
primitive types, such as twos and threes, and, from an 
evolutionary point of view, such precede alternate and spiral 

* I.e. in Exogens. f Bot. Text-Bool, p. 175. % L.c, p. 120. 


On the other hand, whorls of threes, and fives, and others 
in flowers are compressed cycles of spiral arrangements. They 
are, therefore, attempts at simulating ancestral or the verti- 
cillate conditions, but cannot possibly be primitive whorls 
themselves. That the petals can thus become decussating 
with the sepals is a result of the fact that their cords are not 
strictly superposed to and confluent with those of the latter. 
The total number of cords in the pedicel being usually limited 
to the same number as there are parts in the perianth, i.e. 
the calyx and corolla together, there is ample room for them 
to arrange themselves at equal angalar distances around the 
central medulla of the pedicel. Then from the vascular 
cylinder thus formed, they pass off into the sepals and petals 

The sepals and petals or the two whorls of a perianth 
being thus provided for as to their fibro- vascular cords, the 
stamens and carpels, as already stated, generally depend upon 
these latter for their positions, and various arrangements 
arise according as the cords of the perianth-leaves give off new 
members or not. Theoretically there should be at least one 
whorl of stamens superposed to the sepals, another superposed 
to the petals, and two whorls of carpels as well ; but while 
many flowers have both staminal whorls {Gary ophy Usee, 
Leguminosce, Ericacece, etc.), many others, as the Gamopetalce 
retain only one, and more generally the first formed or 
sepaline, but sometimes it is the petaline, as in Primulacece ; 
the probable cause in each case being certain exigencies in 

* That foliar organs possess this power of rearranging themselves 
according to requirements is evident from other considerations ; thus, 
many plants having freely growing erect shoots — as, for example, 
the common Laurel — have their leaf-arrangements represented by the 
fractions | or §, but when extending horizontally, as in the usual con- 
dition, they are distichous. Similar features are seen in the Jerusalem 
Artichoke, which often changes its phyllotaxis on the same stem. 



St. J>e&. 

the flower, througli which nourishment is withdrawn at 
certain places to produce hypertrophy elsewhere. Thus the 
sepaline cord, instead of bearing an anther in Primula, bifur- 
cates at the angle, and each branch proceeds np the margin 
of a lobe of the corolla, and aids in nourishing the latter. 

As a converse instance of the sepaline cord undertaking 
a considerable amount of work, may be mentioned Campanula 
medium. In this plant the 5-lobed fibro-vascular cylinder 
of the pedicel sends off five cords 
intended for the calyx (Fig. 8, sep.) ; 
but, befoi'e reaching the base of the 
superior sepal, it sends off an inner- 
most and lowest cord to become the 
dorsal one of the carpel {d. car.), 
which, in this flower, is thus super- 
posed to a sepal. It also sends off 
two, right and left, one for each 
petal alternating with it (pet.) ; so 
that each petal receives two cords, 
one from each adjacent sepal, — a 
most unusual condition of things, 
for petals have almost invariably 
their own cords issuing from the 
pedicel. Lastly, the same sepaline 
cord provides that of the stamen 
(st.) superposed to it. In this 
flower, therefore, we can understand why there is no petal- 
ine whorl of stamens ; simply because the corolla does not 
possess its own proper fibro-vascular cords to give rise to 

On the other hand, in the Malvacece after the axis has 
supplied cords for the sepals, others furnish those of the 
corolla ; these latter, however, by radial division form two 



Fig. 8. — Vertical and transverse sec- 
tions of the wall of the inferior 
ovary of Campanula medium 
(after Van Tieghem). 


to each petal, subsequently dividing into several; for the 
same pair by repeated tangential division gives rise to the 
series of stamens (which have been thus doubled) superposed 
to each petal, both having arisen from a common cord. 

With regard to the numerous carpels of Hollyhock, I 
find that the axial cylinder which has given rise to the five 
sepals continues on, and by radial division again supplies 
cords to the carpels, which are grouped into five sets super- 
posed to the sepals, as may be easily seen if the pistil be 
examined from below. Hence, as the sepaline or petaline 
cords in these flowers each undertake to form a large number 
of extra parts — many stamens in the one case, and many 
carpels in the other — it is presumable that neither sepaline 
stamens nor petaline carpels could be formed. 

With regard to the presence, and consequently the relative 
position, of one whorl rather than the other of the gynoecium, 
it is due to the fact that sometimes the sepaline cord will 
give rise to the dorsal carpellary, as in Althoea and Campa- 
nula; at others, it is the petaline, as in Fuchsia, Sedum, Ivij, 
etc. ; so that the carpels become superposed to the sepals or 
petals accordingly. As instructive instances of variations in 
this respect occurring in the same family, it may be mentioned 
that all species of Campanula which have five carpels, as also 
Wahlenbergia capensis, Micliauxia, Canarina, and Lightfootia 
suhulata, have their carpels superposed to the sepals and 
stamens. On the other hand, Musschia {Campanula aurea, L.) 
Platycodon (0. grandiflora, Jacq.), and Microcodon have the 
carpels superposed to the petals. 

The fact that either the sepals or the petals can have 
the carpels superposed to them respectively, just as they can 
each have a whorl of stamens, and that, in some few orders, 
the two whorls are actually present, as in Butomeoi and 
Juncagineoi, led me to assume two whorls as the primary 


or ancesti^l number of carpels in an ideally complete 

Besides the usual alternation of whorls resulting from a 
regular and equal displacement of every part of the whorl, 
there may be unequal displacements ; thus, while Cistus has 
a pentamerous flower, with strict alternation of its whorls, 
Helianthemum has a tendency to be trimerous ; first, in the 
two outer sepals being reduced in size, and the pistil to three 
carpels instead of five. In this flower there are five petals, 
but in correlation with the preceding irregularities, it will be 
found that two pairs of petals stand superjDOsed to the sepals, 
Nos. 3 and 5, while a single petal is over No. 4 ; Nos. 1 and 
2, therefore, have none superposed to them. With regard to 
the stamens, it may be added that those of Cistus consist, 
first, of one whorl of five, the most interior and first developed 
superposed to the sepals ; and a second whorl superposed to 
the petals, in which the stamens are grouped into five clusters. 
The staminal whorls arise centrifugally. 

Another cause of a change of order in the whorls results 
from substitution of one kind for another. Thus, in the 
female flower of Z anthoxylon, the five carpels are superposed 
to the five sepals. In the male, five stamens now occupy 
exactly the same place as the carpels, the corolla alternating 
with the sepals in both kinds.* 

The interpretation I would suggest is that the sepals, 
being the only whorl of the perianth developed, the calyx is 
the only source for supplying the dorsal cords of the carpels 
which thus become necessarily superposed to them. 

From what has now been said, it will be seen that the 
arrangement of the essential organs of a flower is, as a general 

* See Figs, in Le Maout and Decaisne's Descriptive and Analytical 
Botany, p. 324. The female flower is described as apetalous, but Payer 
discovered rudiments of the petals. 


rule, most intimately connected with the union of their fibro- 
vascular cords with those of the perianth ; and as parts of 
flowers are often mnltiplied, as the petals of Camellia, 
perianth-leaves of Daffodils, etc., such has given rise to the 
idea of chorisis or dedouhlement of French authors ; as if 
one organ had split into two or more. That vascular cords 
can become repeatedly bifurcated is abundantly observable, 
whether radially, as in the case of the carpels of the Holly- 
hock, or tangentially, as in producing the stamens of the 
same flower. The more correct way, therefore, of regarding 
the process would seem to be, first, to recognize the phyllo- 
tactical origin of the perianth as the basis to start from, and 
then to regard each fibro-vascular cord as an instrument 
for furnishing any number of appendages, whether they be 
additional petals, stamens, or carpels, by the process of 
chorisis, not of the complete organ, as generally meant, but 
of the cord belonging to it. 

To summarize these remarks — we find that the cause of the 
alternation of the whorls of the perianth, or of the calyx and 
corolla, is due to their being made up of cycles of spiral 
arrangements, which are projected on to the same plane, and 
so form verticils. Their positions are then shifted so that 
the parts of each whorl bisect the angles between the parts 
of the whorl succeeding or preceding it. 

Secondly, having laid this foundation, the stamens and 
carpels follow in superposition to one or other or both of 
the preceding whorls in consequence of the branching 
of the fibro-vascular cords. And this accounts for super- 

It may be still further inquired why in some cases the 
sepaline, and why in others it is the petaline cords which 
give rise to a whorl of stamens or carpels, as the case may 
be. The reply at present must be speculative, for there may 


be more than one influence at work to determine what whorl 
shall follow each of those of the perianth. 

The immediate cause is nutrition ; but the deeper question, 
what directs the nutrition to one cord rather than another, 
can only be guessed at in most cases : but as the petaline 
stamens are generally absent from at least the gamopetalce, 
it would seem that the enhancement of the corolla through 
the agency of insects has caused the whorl of stamens in 
front of it to be atrophied through compensation. Some 
special circumstance, however, we know not what, have 
interfered to retain that whorl in Primulacece, and some few 
other plants. 

The reader must be reminded, however, that this method 
of branching in order to give rise to stamens and carpels 
from the cords of the perianth is not universal. When they 
are many, it is done by the fibro-vascular cylinder of the 
pedicel becoming much enlarged, and consisting of a great 
number of cords, all arising by lateral chorisis, it is true, 
but long before they enter the floral members ; so that by 
the time the latter are about to emerge they each receive 
their own cords from the general axial cylinder. This is 
what happens e.g., in Rammculaceoe and Cruciferce. 




CoHESiox. — General Observations. This term signifies the 
iinioTi between parts of the same kind or whorl ; and the 
prefix gamo- is used in conjunction with the terminations 
-sepalons, -petalous, and -phjllous, — to indicate that the parts 
of the calyx, corolla, and perianth respectively cohere. In 
the case of the stamens, they are said to be mon-, di-, tri-, or 
poly-adelphous, according as the filaments cohere into one, 
two, three, or more groups ; while syngenesious is used for 
the coherence of anthers, and, lastly, syncarpous denotes that 
the carpels of a pistil cohere. 

There are two kinds of cohesion, congenital and by con- 
tact.* Congenital cohesion I regard as an advance upon 
freedom, or a further state of differentiation ; for, according 
to the principles of Evolution, freedom or separation of parts 
must precede their union ; just as, for example, bones are 
free in the embryo which become " ankylosed " in the adult ; 
or always free in a fish, while their homologues cohere in 
higher types of vertebrates. 

Congenital cohesion applies to by far the greater number 
of cases of union amongst the parts of the different whorls 

* We might appropriately distinguish these two kinds of union by 
the terms connate or "born together," and coherent or "sticking 


of flowers, respectively. Cohesion by contact is the cause of 
the anthers being sjngenesious in the Compositce. It applies, 
sometimes at least, to the two margins of each carpel when in 
contact np the axis of an ovary, as of that of a Lily. The 
stigmas of Asclepias are at first free, but later in their deve- 
lopment they become coherent by contact. 

Congenital cohesion takes place almost from the very 
commencement of growth and development of the parts, so 
that when full-grown there may be no trace of the line of 
cohesion. Fibro-vascular cords, indeed, often occur in the 
very position of it, not unfrequently branching off in various 
ways, as, e.g., at the fork to nourish the adjacent free portions 
of the limb. This occurs in the calyx of Stachys and the 
corolla of Frimula, etc. In Gavipanula rotundifolia the fibro- 
vascular system of the corolla becomes completely altered, and 
instead of representing that of distinct leaves in contact by 
their edges, the veins ramify and anastomose all over the 
general space between the two adjacent dorsal ribs, com- 
pletely obliterating all trace of the line of union between 
them. In the case of the Primrose, however, the calyx has 
the exact appearance of five pinnately nerved leaves being 
united by their thin and impoverished edges, where there is 
nothing but translucent tissue without any cords at all. 

It is important to observe this more or less complete 
modification of the fibro-vascular system under congenital 
cohesion, as it shows how much more highly differentiated 
a condition has been acquired than when the parts are free. 
In the latter case they represent more closely the forms and 
venation of distinct foliar organs. 

As a curious instance of cohesion of both kinds in the 

same organ, may be mentioned the corolla of Phyteuma ; the 

basal portion of which consists of five petals congenitally 

united ; but the five portions of the limb cohere by contact 



at the apex, and so form a tube which collects the pollen 
shed into it by the five free anthers, which are included 
wdthin this corolla-tube (Fig. 9). 
They thus form the " cylinder " for 
the " piston " action of the pistil which 
continues to grow, and so sweeps out 
the pollen beyond the extremity of 
the tube, just as it does from the 
syngenesious anthers of the Covi- 
positce and Lobelia. The five portions 
of the corolla thus cohering by con- 
tact subsequently become more or less 

The rationale of Cohesion lies in 
its adaptation to insect agency, and 

Fig. 9.-i>7.,ieuma (after Muller). .^^^.^g ^ ^^^^^^^ ^^^^^^ ^^ Spccializa- 

tion than when the parts of the whorls are free. Thus in 
Thalamiflorce, of such an order as Ranunculacece with regular 
flowers and with all the parts of the perianth whorls free, the 
flowers are usually visited by a much greater number and 
variety of insects than are those of orders of Corolliflorce. For 
example, Miiller records sixty-two species of insects as seen by 
him to visit Uanunculus acris ; whereas the humble-bee alone 
enters the gamopetalous tube of the Foxglove. This adapta- 
tion oiform to insect visitors will be better appreciated when 
we come to discuss that principle of Variation, which so 
powerfully affects floral structure. 

It occasionally happens that parts normally united become 
free : the process is called " dialysis," and may be regarded 
as a reversion to an ancestral free condition. Fig, 10 repre- 
sents a flower of Mimulus in this condition. The rationale 
of cohesion in the sepals, petals, and stamens, I regard as the 
immediate result of hypertrophy set up by insect agency, 



Fig. 10. 

-3Iiinulus undergoing "Dialysis' 
(after Baillun). 

aided by the close proximity of the parts ; and as a resulting 

effect, is the ever-increasing adaptation to the requirements 

of insects, which are more and 

more specialized for them, so 

that, for example, Lepidoptera 

are almost solely adapted to 

long tubular flowers like the 


An analogous process of 
congenital cohesion is well seen 
in the fasciation of stems which 
occurs particularly often in 
succulent shoots, as Asparagus, Cabbage, Lettuce, and the 
young shoots of the Ash tree. This is most reasonably 
referred to hypertrophy coupled with the close proximity 
of the buds which ought to have developed into independent 
shoots. Again, cohesion between the sepals or petals of 
Orchids is not uncommon abnormally under cultivation ; and 
would also seem to be due to the stimulatiug conditions under 
which they are artificially cultivated. 

Hypertrophy in an organ is due to a special flow of 
nutriment to it ; and cohesion may result from the close 
proximity of the parts of the whorl to one another ; but the 
influence which brings about the determination of sap to a 
particular point, I take to be the mechanical strains induced 
by the insect visitors when alighting upon the flower in 
search for nectar or pollen. 

If this principle be correct, that the tubular structure of 
calyces and corollas, as we see them now, has arisen through 
the requirements of those organs to meet strains throwm upon 
them; I think it will furnish the solution to many a question 
that may arise as to the peculiar shapes of corollas, etc., 
besides explainiDg the very principle of cohesion itseK. An 


insect alights on one or two petals. In order to support it, 
an immense gain is secured if the flower can call in the aid 
of the other petals ; and this is obviously obtained by their 
cohesion into a tube, just so far as the required strength is 
wanted. Nothing would be gained by the portions of the 
limb being united, as far as additional strength was required 
to bear the burden. The tubular structure is the strongest 
possible, and when short, as in rotate corollas, little extra 
aid is required ; but if it be long and visited by heavy insects 
and not by Lepidoptera, which hover in front of the flower 
and only insert their long and slender proboscides, then the 
tube finds additional support in the calyx being tubular as 
well. At other times mutual support is gained by the close 
contact of the flowers, as in a capitulum of the Comjpositce, 
from which the calyx vanishes. 

Of course, every degree conceivable is met with between 
short, stout, and strong tubes with no additional aid, and 
slender ones supported by a strengthened gamosepalous calyx. 
These are adapted to insects which alight upon the corolla 
limb ; while for Lepidoptera the tube is more elongated, and, 
as no weight is thrown on the anterior petals, no extra 
support is required. That this is the true interpretation of 
the origin of a gamopetalous corolla, appears from such 
negative evidence as is seen, for example, in Lonicera Peri- 
clymenum and Asperula taurma* which have greatly elongated 
and contracted tubes, deriving no support from the arrested 
calyx; and although somewhat two-lipped, the anterior 
member is no larger than the others ; the reverse being 
always the case when a heavy insect is the regular visitor. 
These two species are exclusively fertilised by the Lejjidoptera, 
such as the Hawk-moth, which only hovers in front of the 
orifice, but throws no weight upon the corolla. 

* See Miiller's figures, Fertilisation, etc., pp. 296, 303. 


We may see, as it were, Nature's first attempt to form a 
tubular process in tlie Cruciferce. Here it is obtained by 
simple approximation of the slender claws of tlie petals, 
wliicb are supported by the erect and closely imbricated 
sepals. A step further is gained in Dianthus, in wliich the 
sepals cohere but the petals are still free. The tliird and 
last stage is arrived at when both calyx and corolla are 

Subsequent to this state of cohesion many additional 
structures may arise as they are required in the formation 
of ribs, etc., as already explained ; while the very form of 
the tube may change from a purely straight cylinder to a 
curved or expanded funnel, etc., according as special strains 
have to be met, which the original form was not well calcu- 
lated to sustain. 

These changes of Form will be more fully discussed when 
I treat of that principle of Variation. 




Cohesion of the Sepals, or Gamosepalous Calyx. — This is 
congenital, and may be free, as in the Carnation and Primrose, 
or associated with a " receptacular tube," as in Leguminosce 
and Rosacece. 

As sepals mostly represent the petioles of leaves, the 
tubular part of a gamosepalous calyx consists really of the 
fusion of the expanded petioles, the teeth of the limb being 
all that remains to represent the blades which are usually 
suppressed. The main fibro-vascular cords correspond to 
the mid-ribs, while the interspaces are either without additional 
" marginal " cords, as in the Primrose, or with single or double 
cords in the line of junction, as in the Lahiatce ; or they may 
be covered with anastomozing reticulations without any linear 
cord at all, as in Mimulus. 

With regard to the presence of linear cords in the line of 
suture, if there be five sepals, there will be at least ten ribs 
to the calyx ; i.e., if there be only one marginal cord ; but 
as there are two margins which cohere, they may have a 
separate cord apiece ; and then there may result fifteen cords 
in all. Thus Stachys has five dorsal cords with barely traces 
of five marginal ones ; Ballota has ten, and Nepeta fifteen. 

The above arrangements may be modified by the separa- 
tion of the two marginal cords in certain places but not in 


others, while supernumerary cords can be formed, which 
appear to have for their function to strengthen the calyx to 
meet the strain upon it when an insect alights upon the 

In the calyx of some species of Salvia, which is strongly 
bi-lobecl, though retaining its five teeth, three dorsal (d) 
are posterior and two are anterior. There 
are two sinorle marmnal (m) cords between 
the three posterior and dorsal, which corre- ^ ^ 

spond to the mid-ribs of three sepals. The m m 

two lateral and marginal cords are each 'ni m 

double ; while a supernumerary cord (s) lies '-'' ^ 

beneath the lip of the corolla between the 
two anterior marginals. The accompanying 
diagram of the sepaline cords of S. Verhenaca will illustrate 
the arrangement. 

The arraugement of the cords (m and s) shows that the 
strain being greater on the anterior side, the calyx has, as 
it were, stretched in that direction, the two marginals having 
separated so widely in front, as to require an extra .cord (s). 
The two lateral ones have not separated to so great an extent, 
while on the posterior side, where little or no strain is felt, 
the marginal cords have remained single. 

As the cord (s) shows how Nature can add a fibro-vascular 
cord if required, so one or more can be subtracted by atrophy 
where no stress occurs. Thus the petals of the Compositce 
have no dorsal or median cords, the five sepaline only being 
present below, but pass up the margins of the petals. Con- 
versely, in the Primrose, the calyx, giving no support to the 
corolla, has no marginal cords. 

The above diagram will represent the distribution of the 
sepaline cords of S. glutinosa and other species, as well as 
S. Verhenaca, but in 8. pratensis the strain has apparently 


been not so great, consequently the supernumerary cord (s) 
has not been developed. 

Such slight differences are significant, because they show 
how readily an organ can respond to different degrees of 
force brought to bear upon it by different insect visitors ; 
and the cords are invariably placed just where the strains are 

The number of ribs to the calyx has been adopted by 
systematists as generic characters in some of the Lahiatce, as 
well as the tabular or campanulate shape of it. Now, it will 
be found that the shape corresponds with the requirements 
of the corolla ; so that if the tube of the latter be compara- 
tively short and slender, the calyx completely encloses it, and 
has its surface strengthened by a variable number of ribs 
according to the genus ; though they are not always constant 
on the same plant. As examples, may be mentioned, Mentha 
and Melittis, which have a broad campanulate calyx, and a 
broad tube to the corolla. St achy s has 5-10 ribs surrounding 
the cylindrical corolla-tube. Galeopsis versicolor has 10 
prominent ribs, and 10 others which reach from the base of 
the calyx-tube to about half-way up. Melissa has a very 
narrow elongated calyx, which fits the slender tube of the 
corolla exactly, and has 13 or 14 ribs.* Similarly Nepeta 
Cataria and N. Glechoma support the contracted slender basal 
part of the corolla-tube, and have 15 ribs to the calyx. 

Teucrium Scorodonia has only 5 dorsal ribs and 2 (posterior) 
marginal. Tiie calyx is very broad compared with the slender 
corolla-tube, and scarcely, if at all, supports it. This flower, 
is visited both by bees, and nocturnal Lepidoptera which 
suck without throwing any weight upon the flower. 

Cohesion of Petals, or Gamopetalous Corolla. — As 

* This difference in the mimber of ribs depends upon the lateral and 
marginal being single or double. 


already stated, this is congenital, and, as with the calyx, so 
with the corolla, the line of junction may be marked by a 
marginal cord, or the interspace covered with reticulations as 
in Campanula rotundifolia. 

As in the calyx of many Labiates, so there may be super- 
numerary cords in the corolla, until they may be greatly 
increased in number, as in Convolvulus Sepium, Digitalis, etc. 
The cords being straight in the tube may ramify in the 
lobes, adding thereby marginal veins to the latter, as in 
Trimula and the Compositce. In this last, the petals are 
devoid of median nerves, hence the importance of the mar- 
ginal with their branches up the edges of the corolline lobes. 

It would be superfluous to multiply examples if the 
principle be understood ; and what I particularly wish the 
reader to realize is the, so to say, extraordinary plasticity 
which resides in these organs of flowers, in that they 
evidently have the power of altering their structure to meet 
a variety of requirements ; so that if we might compare them 
to architectural buildings, we might say that the floral 
Architect at one time saw not only a chance of some orna- 
mental improvements in a frieze at some particular place, 
graceful lines of colour or curvature in another ; or, again, 
flutings, depressions, and elevations, etc., all breaking up any 
chance of monotony : but cunningly adds elegant buttresses 
without, as well as runs up ribs of masonry within the 
walls ; which, while intended to meet particular strains, only 
add additional charms to the general and harmonious beauty 
of the entire fabric. 

Cohesion op Stamens — (1) " Adelphous " Filaments. — 
This occurs in various degrees, from a comparatively slight 
union at the base, as in Linum usitatissimum, to a short 
distance from the anthers, as in Malvaceae and Leguminosm. 
It is undoubtedly an adaptation to insect agency. 


If tbe stamens be monadelplious, and the union be extended, 
it may completely enclose the usual honey-secreting surface 
characteristic of allied genera, the result being that it can 
secrete none at all. In such cases, insects are deceived in 
visiting the flower, as in Genista, and some other mona- 
delphous genera of LeguminoscB. Otherwise, the honey is 
secreted by some other source external to the staminal tube, 
as ia Linum catharticum ; in which flower five inconspicuous 
glands occur on a fleshy ring, just opposite the stamens. In 
Malva, the honey is found in five pits between the bases of 
the petals, and in Pelargonium in a long tube formed by one 
sepal, the insertion of which remains far below that of the 
others, which are carried up by the growth of the pedicel. 
In Laburnum* as in Orchis, instead of a secretion, the fluid 
is only to be secured by piercing succulent tissue which is 
found in front of the vexillum in the form of a cellular 

In diadelphous species of the LeguminoscB, the honey may 
be secreted by the inner basal portion of the staminal tube,* 
or else, and perhaps more usually, by an annular disk which 
surrounds the short pedicel of the ovary, as in Pisum. In 
this case the honey is easily secured by the proper insects 
as the superior stamen is free, and there is also an additional 
facility of access by means of an oval space formed by the 
widening of the staminal tube just above their base. 

In Cercis, the disk is very large, and the 10 stamens stand 
in depressions around it. Consequently they are entirely free. 

The staminal tube, together with the petals, which are 
more or less interlocked together, protect the honey from 
being rifled by the wrong insects,f as it can only be secured 

* According to Miiller. 

t A curious additional protection occurs in Mippocrepis comosa, in 
that the claw of the vexillum, which is elevated in a remarkable manner, 


by such as have proboscides of sufficient length to reach it, 
corresponding, of course, to each species or genus. 

Papilionaceous flowers being irregular, and visited in but 
one way, it is only the superior stamen which is free ; but the 
staminal tube is often imitated in other flowers where there 
may be no cohesion at all, as by the tribe Ocimoidece of 
Labiates, ColUnsiahicolor of the Scrophularinece B.nd Poly gala, 
etc. Similarly, in the case of regular flowers, the mona- 
delphous condition may be closely mimicked by filaments 
which are stout and sufficiently rigid to form a column. 
This occurs in CrucifercB, Viola, Convolvulus, Crocus, etc. In 
some cases, as in Cramhe and Deiotzia, the filaments are pro- 
vided with wing-like structures which render the tube more 
complete. In orange flowers, a certain amount of cohesion 
is actually obtained between some of the filaments. 

(2) Syngenesious Anthers. — These, as stated, are not 
congenitally united, but by simple contact. As with fila- 
ments, so with these, it is an adaptation to insect fertilisation. 
Jasione montana furnishes a good instance for an incipient 
stage where they just unite at their bases only. This cohesion 
is completed in the genus SynantJiera of the same order 
Campamdacece, as well as in the sub-order Loheliece. In other 
cases of true syngenesious anthers there is a complete lateral 
fusion, as in Lobelia and Compositce, in Gloxinia and Im- 
patiens. In all these cases the cohesion is by lateral con- 
tact only, and not congenital ; that is to say, the papilljB of 
the future anthers on emerging from the axis grow to a 
somewhat considerable stage of development as incipient 
anthers before coming into contact. They then coalesce, 
apparently by a slight solution of the surface of the cellular 

carries a triangular flap, which exactly covers the orifice leading to the 
honey. A somewhat similar flap occurs in the petals of Phaseolus and 
Delphinium, which likewise keeps out unwelcome guests. 



walls wliicli touch ; so that when they are fully grown the 
cohesion is firmly secured. An imitative cohesion is seen 
in the anthers of the Heartsease, which arises from the 
interlocking of marginal hairs down the sides of the cells. 
Anthers, when thus closely approximate without actual 
cohesion, are usually called " connivent," as in EricacecE, 
and the word is perhaps appropriate to those of Solanum 
Dulcamara; but in this plant the union is very close, and 
might even be considered as syngenesious. 

The rationale of the close approximation of anthers, or of 
actual cohesion between them, is the effect of insect agency, 
just as for the filaments ; but the method of extraction of the 
pollen varies. In Viola, the proboscis is thrust through a small 
orifice between the connectival appendages 
of the lower pair of stamens, in order to 
reach the end of the honey-collecting spur. 
In Heaths and some of their allies, the 
anther- cells are at first in contact, and so 
prevent the pollen from escaping ; but each 
anther is provided with two auricles which 
extend to the corolla. A bee on entering 
first strikes the projecting stigma, but its 
proboscis soon turns one of the auricles 
aside, which, acting as a lever, dislocates 
the rest, and a shower of pollen falls out. 
In Composite and Lobelia there is a true 
piston action. The style continuing to elongate drives the 
pollen out of the cylinder formed by the anthers, and elevates 
it above the flower, thereby rendering it easy to be dispersed 
by insects. This is well seen in Centaurea (Fig. 11) ; {a) 
represents the stamens with the anther-cells closed above by 
the connectival appendages. The arrow shows the direction 
of the insertion of the proboscis of a bee to reach the annular 

Fig. 11.— Stamens of 


honey-disk at tlie base of the style ; in h, the style-arms have 
spread after protrusion through the separated connectives. 
The brush-like tuft of hairs has swept the pollen out by 
means of the piston-action of the style. 

In Campanula, the action is different, for the anthers 
though connivent, have not yet become syngenesious, as in 
allied genera, e.g. Lobelia. They at first closely surround 
the style, which is provided with long collecting hairs upon 
which the pollen is caught. The anthers then shrivel and 
fall down. Subsequently a bee enters the expanded bell, 
grasps the style with her legs, and so transfers the pollen to 
the abdomen. This method is identical with that followed 
by bees in getting honey from Crocus, though in this 
genus the anthers remain erect, and, being extrorse, at once 
discharge the pollen upon the insect without the interven- 
tion of the style. 




Cohesion op Carpels, or Syncarpous Pistil. — The accepted 
doctrine that the carpels are metamorphosed leaves, will be 
considered more fally when teratological modifications 
come to be discussed ; and the proof that an ordinary 
carpel, such as a legume, is merely a leaf folded upon itself 
in a conduplicate manner with the margins coalescing and 
then metamorphosed into a new organ, requires no special 
evidence now. That a syncarpous pistil consists of two or 
more carpellary leaves coalescing is equally admitted; and 
there are two methods of cohesion. Either the carpels may 
be ah initio composed of unclosed leaves, which cohere by their 
edges * respectively in contact, thus forming a single cavity 
provided with parietal placentas, — such a union implying 
a more primitive or arrested condition, from an evolutionary 
point of view ; f or they may be individually more or less 
closed before coalescence takes place, in this case by their 
lateral surfaces. The axile placentation is the result. 'The 

* The theory that the placentas are, at least in part, axial, -will be 
seen to be erroneous in consequence of the orientation of their vascular 
cords {e.g. Fig. 12, c, p. 64; and Fig. 13, a, h, p. 65). 

t Thus the parietal placentation of Orohanche is probably a result 
of degradation through parasitism, from the axile, of the Scrophularinece. 
It may be compared to a " cleft palate " and " hare-lip " in man. 


margins show every degree of union from a mere contact 
without real cohesion, thence, cohesion by contact, to a 
solid central axial structure formed by congenital cohesion. 
Lastly, the ovary may be one-chambered, with a free-central 
placenta, as in Caryophyllece and Primulacece ; or with one or 
more ovules attached at the base, as in Bumex, Compositce, 
Graminece, etc. It is these latter kinds especially which 
have given rise to mucli discussion as to the real nature of 
the placentas, and as to how far the axis .enters into their 
construction. To ascertain this latter point, a study of the 
distribution and structure of the fibro-vascular cords of 
the axis and of the carpels would seem to afford the most 
promising clue to the interpretation. 

It has been already mentioned that the dorsal cords of 
carpels generally arise by lateral division from those of the 
sepals or petals ; and then the carpels will be superposed to 
the one or the other of these organs respectively ; * or, a group 
may emerge from the axial cylinder in a horse-shoe form, as 
seen in section; the outermost cord becoming the dorsal- 
carpellary, and the ends of the curve the marginal. This is 
the case, for example, in Cyclavien. 

The point, then, at which the carpellary cords branch off 
from a common stem in the first case may be regarded as 
marking the termination of their axial character ; and in the 
latter case, at the separation of parts of the " borse-shoes " 
to form groups of threes. With regard to those cords which 
become marginal and placentary, it is important to notice 
the position of their spiral vessels. f If they are situated on 
the side of the cord nearest to the medulla, the cord may 

* See pp. 23, 24, and 42, 43, 44. 

t The cords are, of course, reduced to vessels and soft bast only, the 
former being mostly spiral, but occasionally becoming more or less 
reticulated. I shall adopt the usual word Tracheoe. 


generally be regarded as axial ; if, on the other side, i.e. 
nearest to the ovary-cell, and if transverse sections exhibit 
intermediate positions, in which they are central or scattered 
irregularly within the phloem, they are then marginal and 

They may change their position from one side to the 
other of the cord, as far as I have observed, in three different 
ways. The whole cord may twist to the right or left, as in 
Hellebore (Fig. ,12) ; or, secondly, it may divide into two, 
and each half turn towards an adjacent half of another cord 
andnnite with the latter, as in Pelargonium zonale (Fig. 13, h) ; 
or, thirdly, the tracheae may traverse the phloem and so pass 
ont at the opposite side at a higher level, as in Ivy (Fig. 14,/, 
p. 68). In any case, as soon as the trachea3 are so placed as 
to effect their object of nourishing the ovules, they may be 
pronounced to be unquestionably and strictly carpellary. 

I will take Hellebore as illustrating the first case. Fig. 
12 represents a section of the floral receptacle taken imme- 



Fig. 1 2.— Hellebore : sections at base of ovary. 

diately above the insertion of the innermost stamens. There 
are nine cords* oriented as axial, three of which are beginning 
to curve outwards to form the dorsal cords of the three 
carpels. Sections made a little higher show that the three 
pairs of cords have spread out and revolved so as to bring 
their spiral vessels into a radial direction (&, c). In this 

♦ The tracheae are indicated by black lines or dots, the phloem being 
inclosed within tho thin lines. 



position the tracliese of each pair of cords face each other. 
At this point, then, they have quite lost their strictly axial 
character of facing the centre, and the axis is therefore no 
longer concerned in the structure. A little higher the 
cavities of the ovaries (indicated by the dotted lines) appear 
between the dorsal cord and the pair of marginal ones ; and 
now the latter turn their spirals completely towards the 
ovary cells, having rotated through 90° in all. The object of 
this rotation is to enable them to send off cords to the ovules. 
The second method is well seen in Geranium, Pelargonium 
zonale, and Im-patiens. A section of the receptacle of the 
first two, made between the insertion of the stamens and the 
pistil, shows five groups of three cords each, arranged as in 
Fig. 13, a. Small portions of the ten staminal cords are 

Fig. \Z.— Pelargonium. : sections at base of ovary (a, b, after Van Tieghem), 

seen on the circumference of the section. The outermost 
one of each group of three will form the dorsal cord of the 
carpel. The two inner have their vessels already turned 


towards each other, as described in Hellebore, and are in part 
required for the placentas. They are, therefore, no longei 
oriented as in an axis, i.e. with all the vessels arranged on 
the inner edge of the cord and facing the central medulla. 

A short distance above the base of the pistil, the inner- 
most cords divide in a somewhat irregular manner, but 
rearrange themselves symmetrically round the centre of the 
ground tissue in ten cords, as soon as the ovary cells have put 
in an appearance. The method by which this condition is 
arrived at was described by Van Tieghem in Geranium 
longipes, and with slight modifications it will apply to 
Pelargonium zonale. Each of the lateral cords divides into 
two (Fig. 13, 6), the two interior and adjacent branches 
nnite to form a single marginal cord with the tracheae 
within or on the outer side (Fig. 13, c). The two outermost 
branches pass off to the right and left, and proceed to join 
the corresponding halves from the neighbouring systems. 
The pairs uniting thus form five cords of double origin, 
alternating with the crescent-shaped marginal cords of the 
carpels (r). There are thus formed five in front of the ovary- 
cells, and five in fi'ont of the septa, "which," Yan Tieghem 
observes, " one would regard as axial, if one did not pay 
attention to the mode of formation of the cords and to their 

In his description of Imjpatiens Boyleana, he says that 
the two innermost branches (Fig. 13, h) unite at first end to 
end, i.e. like an 8, with the trachea9 at the extremities in 
contact; they then form one cord with the spiral vessels 
towards the circumference of the section, by rotating through 
90°, accompanied by complete fusion. 

In Pelargonium zonale, the tracheae become plunged, as it 
were, within the phloem-tissue of the cords, as shown in 
Fig 13, c, which then fuse together laterally. 


Above the ovary-cells, at the base and thicker part of the 
style, a section (Fig. 13, d) shows five solid circular buttresses, 
the tissue of which is continuous with the central paren- 
chyma, in the middle of which a lacuna (Z) is formed by 
rupture. In the depression between the buttresses, a small 
portion of the style and conducting tissae forms a bridge, 
as in Fig. 13, d, showing a cavity below it. 

It is in this homogeneous mass of ground tissue that we 
have a complete fusion of the hypertrophied borders of the 
carpels which have thus entirely lost their individuality. 
The axis proper disappeared as soon as the spiral vessels 
became oriented, as in Fig. 13, a. 

Hence the dotted lines radiating from the centre (c) 
mark the ideal boundary of each carpel, and the line across 
the base of the ovary-cell is the place where rupture will 
take place when the fruit is mature. The column, or so- 
called "carpophore," remaining is therefore entirely carpellary 
in its origin. 

The third method by which the tracheae pass from one 
side to the other of a cord is partly seen in the preceding ; 
and 1 suspect that this is the commonest method of all ; for 
though, when axial, the cord has its spiral vessels fixed at 
the inner angle, as soon as a change of position occurs or 
whenever it has to branch, the fixity of the position of the 
tracheae becomes relaxed, and they readily become enveloped 
in the rest of the tissue of the cord, and so pass from one 
side to the other with perfect facility, as will be seen in the 
case of the Ivy. 

When a syncarpous pistil has its ovary inferior — that is, 
imbedded in the receptacular tube — the real state of cohesion 
between the several carpels is masked in consequence of 
their partially undifferentiated state ; the ovaries of which 
then have the appearance of being simply isolated cavities 


sunk within a mass of parenchymatous tissue. In fact, they 
might often be called " falsely syncarpous," a term applied 
to the Pomece^ but which is equally applicable to Ivy and 

In the pedicel of a flower of Ivy, there are, at a distance 
of about three-quarters of an inch from the tapering base of 
the inferior ovary, four fibro-vascular cords (Fig. 14, a). A 

little higher these split 
up into an irregular 
circle (6), and shortly 
above the base of the 
receptaculartube there 
are fifteen (c), ten 
being more towards 
the circumference 
than the other five. 
The outer ten are for 
the sepals and petals. 
The five inner will 
appear superposed to 
the sepals (d), having 
been already separated 
off by radial chorisis 
rather low down; these 
are for the stamens. 
Then from the petal- 
ine cords, by a similar 
method of chorisis, a 
small cord inins up 
the dorsal part of the 
ovary-cell and another up the axis. This fixes the position 
of the five carpels (if so many be present) as superposed to 
the petals (d). There are often only four, or even three, 

Fig. 14.— Ivy : sections from pedicel to summit 
of ovary. 


ovaiy-cells developed. When this is the case, the cords of 
the centre become fused into four or three (2 -f- 2 -f 1) (e), 
and take up a position alternating with the ovary-cells. 
They become even more welded together higher up ; but they 
separate again, to form twice as many as there are ovary- 
cells (/). If there be three, then each cord may bifurcate, 
though they do not all do so in every instance ; so that out of 
12 cords, three ovular cords are given off to nourish the ovules 
(/), and the rest run up the styles, though the total number 
of cords may be less than 12, as variations seem to take place. 

The ground tissue consists of a loose merenchyma, except- 
ing three or four layers of cells below the epidermis, which 
are more compact ; the ovary-cells — seemingly reduced to a 
thickened epidermal layer only — are plunged freely into 
this tissue (e). The cords run up the centre perfectly 
independent of the ovary-cells (e) with their spiral vessels 
on the inside, surrounding a central medulla. Were it not 
for the presence of the dorsal cord, there is nothing to hinder 
one from calling them axial. It is not until they reach the 
top of the ovary-cells that these cords bifurcate and send off 
one branch each into the pendulous ovules, the other branches 
being conveyed upwards into the styles (/). 

The above description will give a fair example of the 
distribution of the cords for supplying the several members 
of the whorls. The reader can estimate how far the central 
cylinder should be called axial. The fact is, that the whole 
of the tissue of the carpels, excepting the thickened internal 
• epidermis covering the ovules, is totally lost in the general 
spongy mass in which they are imbedded. But since the 
petaline cord gives rise to the small dorsal- carp ellary and 
one axial, theoretically these two belong to the carpellary 
leaf ; and on this ground we should feel inclined to regard 
the central cords not as axial but marginal and carpellary, 


notwithstanding tlie fact that the tracheao are oriented 
inwards ; since it is not until they reach the level of the 
insertion of the ovules that they pass either to the middle 
or opposite side of the cord. The rest of the carpellary 
tissues are undifferentiated, as stated above, and it is this very 
common condition in the case of inferior ovaries that has led 
botanists to regard the lower parts of the carpels as being of 
an axial nature and not foliar. 

The Formation of Septa. — With regard to the union of 
the surfaces of the carpels to form the septa, the rule is for 
the adjacent epidermides to be altogether wanting ; and, if 
the median tissue be thick, the walls of two adjacent ovary- 
cells may be very wide asunder, as in the Ivy. On the other 
hand, the septa may be reduced to the two epidermal layers 
alone, and then they are often scarcely coherent at all, as in 
Balsam and Lemon. 

In some cases, the epidermides are not in contact through- 
out their entii^e surfaces, and whenever this is the case the 
characteristic epidermal cells reappear, as in Liliacece and 
AmaryllidacecB. Similarly, as soon as the carpels of Hellebore 
become free, the epidermides of the margins appear in their 
proper character, which now cohere only by contact. It is 
the same with the axile placentas of the Lily. 

As instances where the axis seems to be more decidedly 
prolonged up the centre, are Lychnis and allied members of 
the Silenece. Ph. Van Tieghem has also shown how an axial 
cylinder ascends up the middle of the flower of Campanula 
medium for about two- thirds of the height. Thus Fig. 15, a, 
represents a section of the fluted pedicel ; b shows the lobes 
isolated, each containing a portion of the fibro-vascular 
cylinder. In c, the broken central cylinder has again closed 
up, a section showing a complete cii'cle of an axial character. 
The triangular basal portions of the ovary-cells have now 



appeared, d represents a section of two-tlilrds of the height 
of the inferior ovary; but now the fibro-vascular cylinder 
is dissociated, and forms fifteen separate cords — two being 
marginal to each placenta and one belonging to each septum. 
As the cords have their spiral vessels reversed in position, 
i.e. facing outwards and not inwards towards the centre, 
their axial character has ceased.* 


Fig. 15. — Campanula medium (after Van Tieghem). 

The rule appears to me to be that as soon as, or even 
before the level of the insertion of the ovules is reached, the 
internal position of the tracheaB is abandoned. This is the 
case with Lychnis. 

In some cases there is an apparently axial formation, 
* I do not find that matters can be really expressed quite so "diagram- 
matically " as, e.g., in his fignre d ; for Van Tieghem does not pay much 
attention to the central and scattered positions of the trachece, which I 
take to be quite as significant as their outward orientation ; for as the 
ovules are approached they become dispersed, though a meduUa remains. 


wtich has proved to be misleading. Thus, in Geranium and 
allied genera, the beak-like process from which portions of 
the carpels separate when ripe is not axial at all, but simply 
the coherent placentas of an entirely carpellary origin."^' 
This will be understood from the description I have given of 
Pelargonium (p. 65). 

The raericarps of the fruit of an umbellifer are also 
supported on a carpophore, which is likewise usually described 
as axial ; but anatomical investigations do not warrant the 
conclusion. The commissural surfaces are obviously merely 
the result of rupture between the two carpels which have 
cohered ; and, in consequence of this union, each epidermis 
fails to develop its true character, but remains in an arrested 
condition, having the cells somewhat smaller than the rest 
of the ground tissue. This enables the mericarps to separate 
readily on maturity. A double fibro-vascular cord runs up 
the centre to supply ovular cords at the summit of the 
ovary-cells. If one traces the cords from the pedicel, there 
will be found in the latter a complete fibro-vascular cylinder. 
This spreads out at the base of the inferior ovary into ten 
clearly defined cords which run parallel to each other from 
base to apex, to furnish the petals and stamens ; while two 
only coalesce and form the axial cord. It is this cord which 
constitutes the stylopod when the fruit is ripe. Hence it 
is not axial, but simply the combined marginal cords of the 
two ovary-cells. 

Free Central Placentas.— The position of an ovule or 
ovules on a central support, free from the wall of the ovary, 
or directly on the base of the chamber, and apparently quite 

* Prof. A. Gray (I.e., p. 213) and Henfrey (El Course of Bot., 4tli ed. 
p. 100) both speak of it as axial ; though it was quite correctly described 
and figured by M. Seringe so long ago as 1838 (Mem. sur la Fruit des 
Geraniacdes) : *' Les bords do chaque carpel placentaires sont restes et 
forment la colonne." 



central, has given rise to a good deal of discussion. Two 
views have been taken, one being that such ovules are, in 
some cases at least, axial in their origin, and not carpellarj at 
all ; others would refer all ovules, without exception, to a 
carpellarj source. Analogy, indeed, would, if taken alone, 
seem to justify the latter conclusion, since the numerical 
proportion of ovules having a decidedly carpellary origin is 
unmistakably very great; and any donbt upon the matter 
seems to me to have arisen from a want of due appreciation 
of the arrest of development, or rather failure of a complete 
differentiation Avhich has taken place between the ovary and 
axis at the place where the ovule or ovules appear. 

This arrest is particularly apparent, as already stated, in 
the case of inferior 
ovaries, as of the Ivy. 
Thus, in the Comjiositce, 
the ovular papilla seems 
to arise at the base of a 
cavity in the axis, and 
might easily be thought 
to be axial ; but a slight 
eccentricity may be dis- 
cerned at a certain epoch 
which is the first indica- 
tion of its carpellary 
origin. In Beta the 
basal ovule arises in a 
very similar manner, 
but as the ovary becomes 
more developed, the C 

ovule is carried up so Fig. 16.-5eto (after Payer). 

as finally to become pendulous (Fig. 16, a, h, c.) It is 
much the same in Typha and allied genera. The same 


gradual elevation of the ovule occurs ia Ricinus and other 
Euphorbiaceous plants. 

Similarly, if we compare the differences in allied genera, 
as Bamtnculus and Thalictrum ; in the former genus the 
ovule arises at the very base of the carpel, close to its point 
of attachment to the axis, and remains there. In Clematis 
and Thalictrum^ the marginal cleft of the carpel appears a 
little more decidedly above the base, so that the ovule from 
its earliest period is situated somewhat higher up, and by a 
further development is carried to a yet higher position, and 
so ultimately becomes pendulous. Exactly similar differences 
occur between the orders Compositce and Dipsacece. 

Hence, it would seem that basilar ovules owe their 
positions to corresponding degrees of arrest of the growth 
and development of the carpels, and especially of the basilar 
portions of the carpellary margins. I think, therefore, we 
may draw the following conclusion, that the particular form 
of energy which would cause a carpel to emerge out from 
and be developed freely and entirely from an axis, is more 
or less potential than actual.* Consequently, it develops 
the ovule just where that portion of the carpellary margin 
ivould have appeared had it been formed; so that the tissue 
whence the ovular papilla emerges may be considered to be, 
strictly speaking, neither axial nor carpellary, but undif- 
ferentiated merenchyma, and potentially carpellary. 

From a single ovule we may now pass to pluri-ovular 
ovaries. Dloncea gives us an instance where many ovules 
arise at the base perfectly free from the ovarian wall. In 
this flower the pistil consists of five carpels, which emerge 
congenitally out of the axis, first as a circular rim, which 

* It may be noted that it is more actual in Clematis, etc., in that 
eeveral ovular papillae are produced in genera with pendulous ovules, 
besides being more elevated in position ; but only one in Ranunculus. 


then becomes a cup, wliicTi finally contracts above to 
form the style, just as in Primulacece. lb is, therefore, 
unilocular, while a circle of ovules appears on a thick ring 
of tissue within the base of the ovary. Other circles of 
ovules appear concentrically and centrifugally. It might be 
questioned, therefore, whether the ring which carries them 
were axial or not. I think, however, the same interpreta- 
tion wdll apply here as elsewhere ; that is to say, the ovules 
arise from the place where the bases of the carpels would 
have appeared had they been differentiated out of the axis. 

In the allied genus Drosera the placentas are strictly 
parietal, and the ovules, commencing to emerge half-way up 
the wall, appear successively, both upw^ards and downw^ards. 
Now, as they are centrifugal in Dioncea (corresponding to 
the upivard development in Drosera), it looks as if only a 
portion of the upper half of the carpels were really repre- 
sented at all. 

In this genus there is a barren central space within the 
ringof ovules, perhaps representing the termination of the axis. 

That the basal portion only of syncarpous pistils should 
bear ovules is common enough, and the pla- 
centas often swell out there to form bosses 
which we may reasonably conceive as coa- 
lescing to form the continuous ring character- 
istic of Dioncea. Thus^ce?- illustrates how each 
of the two carpels gives rise to tw^o globular 
protuberances on which the ovules are borne 
(Fig. 17). Anemiopsis, as figured by Payer, 
has a confluent protuberance bearing several 
basifue^al ovules. Similar multiovular bosses Fig. i7— Carpels of 

. cr 7 Ad 7 7- • • ^cer (after Payer). 

occur m bolanece and bcrop/iutarinece, giving 
the characteristic dumb-bell shape in a transverse section. 
Now, if we imagine these swollen ovuliferous placentas 


arising from the basal portions of the carpellarj leaves to 
reach the centre of the ovarian chamber, and be there fused 
together into a solid mass, we should obtain the apparently 
axial structure of Primulacece, Sanfalacecey etc., with the few 
or numerous ovules basipetal in order of development, cor- 
responding to the centrifugal order in Dioncea and the 
ascending order in Drosera. 

The probability that this is the correct view is supported 
by a case I have met with in which the carpels 
of Primula sinensis were dissociated, and more 
or less foliaceons with rudimentary ovules, not 
only along the margins, but with several borne 
on heel-like processes,* which extended towards 
the centre of the ovary, as represented in Fig. 18. 
Anatomical investigations entirely corrobo- 
j^j '^7^ rate the carpellary nature of the central placenta 
of Primulacece. The circle of cords, usually ten 
in number, which pass up the column to nourish the ovules 
are originally separated from the sides of the sepaline by 
radial chorisis, and become superposed to the sepals ; the 
dorsal cords (about ten) having also parted company from 
the five sepaline and five petaline. The latter, however, do 
not give rise to any placentary cords ; hence there are really 
five carpels superposed to the sepals. 

With regard to the position of the spiral vessels, they 
are not oriented as if axial, but are completely embedded 
in the phloem, and consequently central. Moreover, the 
cords in section are circular in form, and not wedge-shaped. 
The central (if not external) position of the tracheae and the 
circular form of the cords are both eminently characteristic 

* Van Tieghera, though once regarding the central placenta as axial 
(Recherches sur la Structure du Pistil, 1868), has more recently arrived 
at the same conclusion as myself {Traite'de Bot., 1884). 



features when they first cease to be axial and become appen- 
dicular. The accompanying diagrams (Fig. 19), (a) Lysi- 
macMa nemorum and (h) Primula verisj will illustrate these 

Fig. 19.— a, Lysimachia nemorum; b. Primula veris. 

remarks. The sections are taken on planes * where the 
pistil is emerging from the receptacle ; s. represents the 
sepaline cords ; ab. st. abortive staminal cords ; p. the petal- 
ine and staminal (combined) ; d.c. dorsal carpellary ; pi. c. 
placentary cords. 

A free central placenta may result from the destruction 
of the septa of an originally axile placenta, as occurs in the 
Caryophyllece. Thus, the ten rows of ovules in Lychnis 
sufficiently indicated tbeir marginal origin. I may add that 
a careful investigation into the origin and distribution of the 
cords has convinced me that the axis in flowers of the 
GaryopTiyllece early ceases to take any part in the structure 
of the pistil. 

* Fig. a represents a section taken rather lower down than in Fig. 6 j 
as the cords in the latter are still undifferentiated in Fig. a. 




Adhesion of Organs. — This term is distinguislied from 
cohesion by limiting its application to the union of different 
whorls. Thus, if the petals or stamens be united to the 
calyx, they are called episepalous, a term usually syrony- 
mous with perigynous ; and if the stamens be adherent to 
the perianth or corolla, they are epiphyllous or epipetalous 
respectively, sometimes also described as perigynous. On 
the other hand, if the stamens and pistil be in close con- 
junction, showing an adhesion between the filament and the 
style, so that the anther and stigma are brought together, 
the term gynandrous is applied to them. 

Adhesion may be safely regarded as an advance upon 
cohesion; and there is, I think, a great probability of its 
being — perhaps, originally, in most if not all cases — a result 
of adaptation to insect agency. 

With regard to the perigynous condition which involves 
a more or less degi'ee of adhesion of the petals and stamens 
to the calyx, this is in many clearly a result of the develop- 
ment of the receptacular tube with its honey-disk lining it, 
as in Eosacece. This causes the free portions of the petals 
and stamens to be carried away from the central axis, and 
placed in a ring "around the pistil," i.e. perigynous ; while 
the more or less amount of adhesion of them to the calyx 


has suggested the term episepalous. In the Rose, however, 
which secretes no honej, the sepals are almost, if not 
entirely free, and articulate readily ; whereas, in other 
rosaceous plants, if the receptacular tube does not itself fall 
off, as in Primus, the calyx remains persistent. 

Although it is usual to regard perigynous petals and 
stamens as episepalous as well — that is, " upon the sepals " 
— when the receptacular tube is well pronouDced, it is more 
strictly in accordance with anatomical structure to regard 
the former as brought into close proximity to the calyx, 
rather than being really inserted upon it. In many other 
cases, as in Lythrum and Daphne, the whole of the tube has 
all the appearance of being truly calycine and not recepta- 
cular; so that "episepalous" will then best describe their 
condition of adhesion. 

It is rare to find a gamopetalous corolla adhering to the 
calyx, but it is so in Cucurhitacece, as in the genera Cucumis 
and Bryonia, where the two outer whorls are united. 

Ph. Van Tieghem observes* that the union may be the 
result of the fusion of the respective parenchymas alone, 
leaving the cords proper to each organ distinct. I think, 
however, that it will be found to be more frequently the 
case that when the cords are superposed, they are fused 
together below, but separate w^hen the organs become free. 
This is well seen in Fninus. The sepaline and petaline 
cords branch, by tangential chorisis, about half-way up the 
receptacular tube, and thus give rise to ten stamens. Each of 
the petaline cords branches on either side again, at a different 
level, by radial fission, and gives rise to ten more.f So that if 
we retain the term " episepalous " for the stamens, we must 
understand that, while the actual stamen is practically free 

* Traite Botanique, p. 390. 

t This will be described more fully below (see Fig. 28, p. 95). 


from the caljx, jet its cord is common with that of the latter 

The epiphjllous or epipetalous condition of the stamens Is 
almost invariably associated with a state of cohesion of the 
perianth-leaves and petals of the corolla ; as exceptional 
instances are Scilla and Lychyiis, which have the parts of the 
perianth and corolla free, but with the stamens adherent to 
them ; while, conversely, Campanulacece and Ericacece have 
gamopetalous corollas, but the stamens not adherent to 

The rationale is primarily, in many, perhaps in every 
case, an adaptation to inject agency. In the majority of 
gamopetalous corollas, the honey usually lies somewhere 
between the insertion of the corolla and pistil, being secreted 
by one or more glands or an annular disk round the base of 
the ovary. There are two positions in which the anthers 
may be placed in regular gamopetalous flowers with reference 
to the visits of insects for the honey; either around the tube, 
as in the Primrose and Scilla, or close around the style, as 
in Convolvulus, Caynpaniila, and Crocus. In the former case, 
when an insect passes its head or proboscis down the tube, it 
touches the anthers on one side of it and the stigma on the 
other ; but as the proboscis may pass on either side of the pistil 
in the same and different flowers, that is on the near or 
remote side, with reference to the position of the insect, such 
flowers have every facility of being crossed. If they be 
heterostyled, as the Primrose, then of course each kind has the 
greater chance of being crossed by the other sort. 

* The distribution of the cords in the floral receptacle of Azalea, 
between the insertion of the corolla and pistil, is \^^y anomalous, having 
no symmetrical arrangement around the centre ; while the cords of the 
corolla of Campanula, as described above, are peculiar for other reasons. 
This may, perhaps, have something to do with the exceptional freedom 
of the stamens from the corolla. 


Iq the case of Crocus, Convolvulus, and other flowers with 
a contracted base to the corolla or perianth, the anthers are 
situated close round the style. In these flowers, the insect 
alights on the stigmas, as already described, grasps the central 
column and sucks the honey head downwards, and so gets 
dusted on the abdomen, the pollen from which is thus trans- 
ferred to the next flower visited. 

The adhesion of the stamens to the corolla or perianth 
thus seems to give a rigidity and firmness, as well as leverage 
in some cases, so that the action of the insects is more 
accurately secured, and some one particular spot on their 
bodies invariably struck and dusted with pollen ; which 
would scarcely be the case if the filaments were free and at 
liberty to oscillate or swing about in any direction. 

In many flowers with irregular corollas, the stamens are 
declinate ; and their adhesion to the tube is then of manifest 
advantage, for the basal part of the filaments thus acquires an 
additional strength to act as a fulcrum, which enables the 
filaments to support the weight of the insect. In JEchium, for 
example (Fig. 20, p. 82), the corolla is even strengthened by 
a rib where the stamen is inserted. This part constitutes the 
fulcrum. The line of force from the fulcrum intersects a line 
perpendicular to the filaments, corresponding to the weight of 
the insect ; while the third and upward force is that exerted 
by the filaments to counteract the resultant of the two former.* 

The origin of the adhesion between the stamens and the 
outer whorls is revealed by anatomical investigations ; for 
the rule is, as described in the case of Frunus, that the fibro- 
v^scular cords of the stamens arise by division from those of 
the outer whorls whenever they are superposed to them. 

In other words, Avhen adhesions are seen between the 
floral whorls, by being superposed to one another, then a 

* See also Figs. 38, 39, and 40, pp. 124-126, and consult text. 



fusion of thefr respective cords will be found. Tf the members 
arise freely, as in UanuncalacecB and Cruciferce, tben their cords 
are inserted into the axis, having arisen by radial division 
or lateral chorisis. 

In the case of the gynandrons pistil, the stamens have 
their fibro- vascular cords more or less imbedded in the recep- 

Fig. 20. — Echium ; a, side view ; b, before, and c, a'ter shedding pollen ; showing 

tacular tube, or rather the common tissue resulting from the 
fusion of the ovary and the tube together ; the anther then 
stands on the summit, and if there be a short or no style, but 
only the stigmas terminating the ovary, then the anther is in 
close contact with it, as in Hippuris, Orchis, etc. When there 
is a style, the filament may be prolonged in adhesion with it, 
as in most orchids possessing the so-called column. It is not 


SO, however, in Aristolocliia, according to Van Tiegliem, 
though often described as such.* 

To summarize the above remarks, it seems clear that all 
adhesions between the two whorls of the perianth, to be 
found mostly in the Calyciflorce, is an accidental occurrence 
due to the hypertrophied condition of the axis in forming a 
receptacular tube ; so that the term " perigynous " is more 
strictly applicable than " episepalous." 

Adhesions between the filaments and corolla, or calyx if 
the former be wanting as in Daphne, is an adaptation to insect 
fertilisation ; whereby a more rigid position is acquired for 
the stamens, coupled with a gain of leverage, etc. 

Lastly, adhesions between the stamens and pistil only 
occur where there is a receptacular tube, or "disk," as in 
Nymplicea ; and the fusion of filaments with the style, or 
between anthers and stigmas, is brought about by the very 
close proximity of the organs when in an early and undif- 
ferentiated state. 

* Duchartre, Elem. de Bot., p. 648; Henfrey, Lc, p. 125 ; Benth. and 
Hooker, Gen. PL, vol. iii., pt. 1, p. 123 ; Van Tieghem, Traite de BoL, i., 
p. 422. 

Van Tieghem's description and figure (Fig. 21) is as follows : — 
" The styles and stigmas are abortive, and the six carpels 
are reduced to their ovaries. It is, then, the thickened 
connectives of the anthers, coherent laterally into a tube 
and covered above with stigmatic papillae, which now play 
the part of stigmas and of the style." 

To judge from Payer's figures (Organog^nie, pi. 91 and 

pi. 109), the stigmas appear to rise from the inner side of the 

very short filaments, and might be interpreted as truly car- 

pellary stigmas, but fused to the former. A further investi- 

gation of the distribution of the fibro-vascular cords should Fig. 21.— Aris- 

be made. Moreover, Asarum does not appear to have anv- ^?'^^li?* (after 
, , . , - ^^ •' Van 1 legnem ). 

thing so abnormal. 




Having now noticed the difEerent kinds of unions, we may- 
ask what has brought them about. 

We have seen how progressively complex conditions can 
be traced from entire freedom, as in Buttercups, through 
forms of Cohesion, such as the gamosepalous, gamopetalous, 
monadelphous conditions, etc. •, to cases of Adhesion, as of the 
perigynous and epipetalous states ; and, lastly, to the adhesion 
of the ovary to the receptacular tube. 

As stated above, these conditions are correlated with 
greater and progressive differentiations of the floral organs, 
which have been brought about by insect agencies. The 
above-mentioned and other terms do not, however, explain 
how or what the immediate influences are which induce 
unions of various kinds amongst the parts of flowers ; but 
some researches of Mr. Meehan on the Coniferce * will perhaps 
give us a clue. There is a well-known and a very generally 
prevailing feature amongst certain genera of Conifers — as of 
the Cupressinece, for example — that the foliage can appear 
under two forms, the leaves being either free from their bases, 
or more or less adherent to the axis. The two forms of leaves 
have been recognized as specific characters in Juniperus, 

* On the Leaves of the Coniferce, Proc. of the American Association 
for the Advancement of Science, 1869, p. 317. 


Betinospora, etc. ; but both kinds of foliage -not infrequently 
appear together on the same plant; and, when this is the 
case, the spinescent and free leaves are borne on relatively 
less vigorous branches, the adherent foliage being charac- 
teristic of the more vigorous and quick-growing terminal 
shoots. It has been also noticed by Dr. M. T. Masters that not 
only do the broad and free leaves of Juniperus and Betinospora 
not occur on the leader shoots, but when the plant is varie- 
gated then free leaves (on the stem with arrested growth) 
are much more variegated than they are on the quick-grow- 
ing leader shoot.* The last-mentioned observer has also 
noticed that the free foliage is characteristic of the younger 
condition of the plant, the adnate foliage that of the adult 

The conclusions arrived at by Mr. Meehan are as follows : 
(1) The true leaves of ConifercB are usually adnate with the 
branches. (2) Adnation is in proportion to vigour in the 
genus, species, or in the individuals of the same species, or 
branches of the same individual. (3) Many so-called dis- 
tinct species of Coniferm are the same, but with their leaves 
in various states of adnation. 

Another very common form of adhesion, to which I have 
already alluded and which is most probably due to hyper- 
trophy through succulency at an early stage, is fasciation.f 
Under this condition the fibro-vascular cylinder of at least 
two " axes," which would be normally separate, coalesce, 
and form an oval cylinder with, it may be, only a slight 

♦ Gard. Chron., 1883, vol. xix., p. 657. 

t For remarks on this phenomenon the reader is referred to Dr. 
Masters's Teratologxj. It is particularly common in herbaceous plants, as 
Lettuces, Asparagus, etc., and not unfrequent in Ash-trees. I observed 
a trailing plant of Cotoneaster growing over a rockery by the side of a 
stream in a garden, almost every branch of which was fasciated. 



constriction indicating the union. The medullas, cortical 
and epidermal layers, are also continuous throughout and 
common to the whole. 

Now, the union of two opposite " appendages " to an axis, 
as in the case of connate leaves, may take place. This may 
be called foliar fasciation in which the fibro-vascular cords 
of each " leaf " are embedded in a common parenchyma, and 
all encased together within a common epidermis. 

If we regard the receptacular tube of, say. Fuchsia and 
Narcissus in the same light, though adherent to the ovary 
like a decurrent leaf of a thistle or Sedum, I see no argument 
against the supposition that the tube, in such cases as these, 
may be regarded as the fasciated petioles of the sepaline 
and perianthial leaves, now adherent to the ovary within 

A pear would seem to combine both axis and petioles, as 
the base of the ovaries is situated much above the commence- 
ment of the expansion of the pedicel (see Fig. 22, p. 90, 
and Fig. 26, p. 94, and consult text). 

Each case must, however, be interpreted on its own 
merits ; and I think there will be little difficulty about this, 
if we recognize the fact that both the pedicel acd floral 
receptacle on the one hand, and the petioles or their floral 
equivalents on the other, can alike assume all the features of 
the so-called receptacular tube. 

Now let us apply these principles of union through 
hypertrophy to flowers, and we have an interpretation 
according to the theory advanced in this book : that differ- 
ences of floral structure depend largely upon different dis- 
tributions of nutrition in the sevei^al organs ; and that the 
irritation set up by insects themselves is one of the most 
potent causes of a flow of sap to certain definite places, 
which encourages local growths, thereby inducing these 


unions to take place between tlie parts of any whorl, form- 
ing "cohesions," and also between different whorls, or 

Other causes may determine them, for hypertrophy may 
set in through a purely vegetative stimulus ; for it is not 
unfrequent to see abnormal cohesions and adhesions in cul- 
tivated orchids, such as petals or sepals adhering to the 
column, etc. Such may, with a good deal of probability, 
be referred to the artificially stimulated conditions under 
which they are grown. These abnormal cohesions between 
members of the perianth, and adhesions to the column, have 
been observed both in this country and America.* As a 
particular instance of the latter kind, Mr. Meehan had 
observed several dozens of flowers of Phaius grandiflorus which 
had the dorsal sepal united to the column, all being confined 
to separate spikes from those which have perfect flowers. 
In some cases, of the same plant two of the petals were 
united so as to form a hood over the column. 

Another peculiarity of Orchids is the tendency to convert 
sepals or petals into labella, and to multiply the spurs when 
an orchid is characterized by them so as to render them 
peloric, a sure sign of hypertrophy. f 

All these "monstrosities" seem to point to an excessively 
unstable condition of equilibrium in the flowers of Orchids ; 
and that they are peculiarly sensitive to the effects of nutri- 
tive stimuli, whether brought about by visits of insects or 
by artificial cultivation. So that the order Orchidece is 
particularly interesting, as furnishing indirect or even direct 

* As by Mr. T. Meelian. Proc. Acad. Nat. Soc. Phil, 1873, pp. 
205, 276. 

t The remarkable influence of the presence of a "plant-bug," 
causing the normally irregular corolla of Clerodendron to become 
hypertrophied and peloric, will be described hereafter (p. 130). 


proof for my theory — that the forms and structures of flowers 
are the direct outcome of the responsive power of protoplasm 
to external stimuli.* 

* We maj, perhaps, see some analogy between these unions amongst 
floral organs, which thus occur abnormally in orchids and normally in 
BO many flowers, and inflammatory adhesions in the human subject. 
It is well known that certain, otherwise abnormal, unions may be con. 
genital, which usually only occur through inflammation set up by 
abnormal excitation, but they are not hereditary. 

I have alluded to hypertrophy and atrophy as causes of the struc- 
tures of flowers, and shall have more to say about them. I would here 
add the following analogous phenomena between the animal and vege- 
table kingdoms. Sir James Paget remarks : — " Constant extra-pressure 
on a part always appears to produce atrophy and absorption ; occasional 
pressure may, and usually does, produce hypertrophy and thickening. 
All the thickenings of the cuticle are the consequences of occasional 
pressure ; as the pressure of shoes in occasional walking, of tools occa- 
sionally used with the hand, and the like : for it seems a necessary con- 
dition for hypertrophy, in most parts, that they should enjoy intervals 
in which their nutrition may go on actively " {Led. on Surg. Path., i. , 
p. 89). 

The reader will perceive the significance of this passage when 
recalling the fact that insects' visits are intermittent. 

Atrophy by pressure and absoi*ption is seen in the growth of embryos ; 
while the constant pressure of a ligature arrests all growth at the 
constricted place. On the other hand, it would seem to be the persistent 
contact which causes a climber to thicken (see p. 156). 



The Calyx or Recefi'acular Tube. — This organ consists of 
a cellular sheath of varying degrees of thickness, free from 
or adherent to the ovary. Much discussion has arisen as to 
the true nature of it, whether it should be regarded as axial 
or foliar. The older view generally maintained was that it 
consisted of the lower part of the outermost whorl of the 
perianth or calyx — in other words, that the basal or petiolar 
portions of the sepaline leaves were coherent ; and if the 
ovary were inferior, then they were supposed to be adherent 
to the latter as well. 

Schleiden appears to have been the first botanist who 
propounded the view that it was axial and not foliar. He 
was followed by others ; but this idea took two forms. 
According to one, it was thought that everything below the 
summit of the inferior ovary — that is to say, the outer wall, 
the septa and placentas — was axial, and only the free portion 
of the summit of the ovary, together with the styles and 
stigmas, were foliar. According to the other view, it was 
maintained that the ovaries, styles, and stigmas were foliar, 
and the superficial covering to the ovary alone was axial. 
The first view was held by Schleiden, A. de Saint Hilaire, 
Trecul, Payer, Prantl, and Sachs ; * the latter by Decaisne, 

* E.g. Sachs' Text-Book of Botany, Eng. (2nd) ed., p. 566. 



Naudin, Pli. Yan Tieghem, and, I think, English botanists in 

There are three methods of investigation, which conjointly 
may guide us to the discovery of the real nature of the tube. 
The first is that of following its development; the second 
is teratological, and the third anatomical. 

Morphological Investigations. — -In tracing the morpho- 
logical development of flowers of the JRosacece, where the 
receptacular tube is a characteristic feature, one notices how 
a border, surrounding the domelike termination of the axis 
which soon produces carpellary papillae, rises upwards and 
elevates the sepals and the papillae of the petals and stamens. 
This border ultimately forms the tube ; and the question is, 
whether it should be regarded as the basal part of the calyx 
or a development from the axis. 

In the Pomece we find the apocarpous condition of the 
pistil, characteristic of all the other members of the Rosaceee 
still retained at first ; but in consequence of the gi^owth and 

close proximity of the tube 
with the carpels, various 
degrees of adhesion are 
brought about between 
them ; thus, in Pyrus (Fig. 
22, a), the bases only of 
the carpels are from the 
first fused into the axis. 
In Cotoneaster (b) the fusion 

ms.22.-a,Fyrus;b,Cotoneaster(nnerr.yer).^^^^^^^ ^^ ^ ^.^^^^ j^^.^j 

on the ovaries. Such " half- inferior " ovaries occur in 
other genera, as Saxifraga granulata, Gloxinia, etc. From 
such we pass to completely inferior states, as in Comjpositce 

* Bentham and Hooker describe the inferior ovary of the Pomece 
in the terms, " Calycis tubus ovario adnatus." 


and TJ'mhelUfe7ce, while Onagracece furnisli illustrations of 
an extension of the receptacular tnbe to considerable distances 
beyond the summit of the ovary, as in Ciixcea, and probably 
Fuchsia and (Enothera are similar cases. A like prolongation 
is seen in some Compositce with "stipitate" pappus, as the 
Dandelion, Tragopogon, Hypochceris, etc. 

In tracing the development of the inferior ovary of the 
Compositce, the cavity of the ovary appears to be sunk below 
the level of the first emergence of the corolla and stamens ; 
and it is this which has suggested the view that the ovary 
is part of the axis, and that only the style and upper portion 
of the ovary which is exposed is foliar. 

On the other hand, since there are abundant cases of 
transitional conditions ; as, for example, between species of 
Saxifrage, — S* umhrosa having an entirely superior ovary ; 
S. granulata, one that is half-superior, and S. tridactylites, 
a completely inferior ovary ; and moreover, if we compare 
the Pomece with the other tribes of Rosacece, comparative 
morphology does not tend to favour the above view held by 
Sachs, but rather inclines one to the impression that the basal 
part of the ovary must be carpellary and not axial, though 
there may be no visible line of demarcation between the 
cauline and foliar structures.* 

The existence of the above-mentioned facts, and many 
cases of reversion to entire freedom by " solution," supply 
good reasons for believing that the development of the 
carpels is moTe or less arrested below, vrherever they are in 
contact with the receptacular tube ; yet they retain their 
power of developing at least one ovule, as is often the case in 

* To regard the septa of an inferior ovary " as the prolongations of 
the margins of the carpels downwards on the inside of the ovary " 
(Sachs' Text-Booh, p. 567), seems to be a very strained interpretation in 
order to fit the axial theory. 


gamopetalous epigynous orders. Moreover, the ovule is not 
strictly basilar and central, but is really situated laterally. 
Anatomical investigations, as we shall see presently, entirely 
confirm this view. 

Teratological Investigations. — Teratological evidence of 
the axial, or in some cases, perhaps, petiolar nature of the 
so-called receptacular tube is tolerably abundant. Thus, in 
monstrous forms of flowers normally possessing inferior 
ovaries, the pistil is sometimes completely arrested, when the 
latter is replaced by a long pedicel which is usually wanting 
or else is very short, as in Honeysuckle, Ejpilohium^ Orchis^ 
etc. (Fig. 23).* Pears not unfrequently furnish similar 
instances, as in the case of the so-called "Bishop*s 
Thumb Pear, which sometimes occurs of an 
elongated form, destitute of core and seeds. 
These fruits, which are merely swellings of the 
flower-stalk, are produced from the second crop 
of blossoms, which have not energy enough to 
produce carpels (core) with ovules or ripe 
seeds." f There is little doubt that the recepta- 
cular tube is, in these cases, converted into the 
ifoiuo^maifurmed I'ot^^ik® structures in consequence of the total 
absence of the carpels from within it. In other 
words, it is axial. 

There are other indications of the tube being axial in its 
nature rather than foliar ; thus, it frequently becomes " pro- 
liferous ; " that is to say, flowers, or even branches, may grow 
out of it, as is often the case with Roses, Prickly Pear, 
Umbelliferce, etc. J Again, certain kinds of Pears, Medlars, 

* a is the interior of the flower, consisting of a cup-like depression 
with two anthers. 

t Gardener's Chronicle, Oct. 9, 1886, p. 464. 
X Teratology, p. 100, seq. 



Roses (Fig. 24), etc., occasionally bear foliage on the 
external surface of the tube, and when the calyx of the Rose 

becomes abnormally folia- 
ceous, stipules (Fig. 24, sf.) 

(V /I I //) y^'^~^'~y ^^y appear at the summit 
^^ ' ' ^'^'^^^—^^^^^^^ of the tube, indicating- that 

Fig. 24. 

-Leaf-bearins? recpptacular tube of Rose 
(alter :\Iasters). 

Fig. 25.— Hawthorn with super- 
numerary free carpils (after 

point to be the base of the sepal. Sometimes supernumerary 
carpels are borne freely on the top, as in the Hawthorn 
(Fig. 25). 

On the other hand, a tendency to hypertrophy is some- 
times discovered in the petioles of leaves of Apples* and 
Pears (Fig. 26, p. 94) ; and a not infrequent monstrosity is 
seen in Fuchsias, where one or more of the sepals become 
foliaceous, and then their petioles are formed but often 
remain more or adherent to the ovary if present, which 
seems to imply that the tube in this plant might be formed 

* Mr. Meehan describes a similar instance of an Apple-tree which, 
never bore flowers but always had an abundance o^ fruit. The latter, how= 
ever, were composed of metamorphosed and fleshy floral whorls. He adds, 
however, that cork-cells were formed abundantly on the outside of the 
apples ; remarking, " It would seem, therefore, that with the lack of 
development in the inner series of whorls necessary to the perfect frnit, 
those which remained were hable to take on somewhat the character of 
bark structure " (Proc. Acad, ^^at. Sc. Phil, 1873, p. dd). 



by, or at least is homologous with, the petiolar portion of 

the calycine leaves (Fig. 

Phj^llomes, however, 
are after all but modified 
portions of caulomes, and 
petioles are still less de- 
partures than are blades 
from the nature of an 
axis ; so that while in 
some cases one is inclined 
to regard the tube as 
more strictly axial, in 
others it seem to be more 
homologous with a sort of 
fasciation of petioles. 

We shall see directly 
that the receptacular tube 
of Prumcs contains the 
basal portions of the cords 
proper to the calyx and 
corolla, so that we might 
regard the latter as, on 
the one hand, axial cords 
preparatory to forming 
the perianth ; or, on the 
other, perianthial cords 
not yet differentiated into 

Similarly, in the case 
of monocotyledonous 
flowers, as the Daffodil, 
since petioles are less dif- 

Fig. 26.— Pear with hypertrophied and sub-fasciate 

Fig. 27.— Fuchsia with foUaceons sepals and 
petals (after Masters). 



ferentiated from blades in this class than in Dicotyledons, 
the inferior ovarj may be due to the combination of the pistil 
with the united sheath-like portion of the perianth, which is 
prolonged above the summit of the ovary just as it is in 
Fuchsia, though it is not so prolonged in the Snowdrop. 

Anatomy of the Receptacular Tube.— Tracing the course 
of the fibro-vascular cords from the pedicel below the flower, 
say of Prunus Lauro-cerasus, the common laurel, there will be 
found to be ten, corresponding to the sepals and petals. 
The cortical tissue and epidermis are continuous throughout, 
from the pedicel to the summit of the tube. It is well seen 
also in the tapering end of a pear, from which the cortex 
gradually widens, while the fibro-vascular cords run verti- 
cally up the middle. Before the cords arrive at the border 
of the free tube of the Laurel, they have given rise to the 
staminal cords by chorisis, as shown in Fig. 28, a, h. Fig. a 

sti sL-j^iJ 



Fig 28. — Receptacular tube of Prunus (after Van Tieghem). 

represents a section near the edge of the tube in which both 
the sepaline (s) and the petaline (^) have given rise by tan- 
gential chorisis to a whorl of stamens {st. 1) ; but the petaline 
by radial chorisis to another whorl {st. 2), i.e. to twenty 
stamens in all. Fig. h represents a vertical view of the same.* 

* The single carpel is represented in Fig. a to show the position of 
its three cords, one being dorsal, and the other two marginal. 


As lon^ as tlie cords are simple, i.e. up to the horizontal 
lines in Fig. 6, there is nothing to distinguish them from 
cords of an axis, as in the pedicel. If, therefore, we regard 
the branches above those levels as belonging to the floral 
whorls, then the " axis " would terminate at different 
heights up the receptacular tube — which would seem to be 
rather too forced a view to be acceptable. 

Hence it would seem preferable to regard it entirely as 
axial until the portions of the perianth issue freely from the 
upper part of it. We might compare these branches of the 
fibro-vascular cords embedded in the axis to those belonging 
to ordinary leaves, which traverse the stem for various 
distances downwards till they ultimately vanish ; only in 
the case of leaves they are not coherent into a common 
cord below, but remain free from each other. Moreover, 
other members of the Bosacece show that they cannot be 
always petiolar ; because in the rose the sepals reveal their 
foliaceous character, first by always bearing rudimentary 
leaflets, and sometimes stipules as well at the top of the 
tube (Fig. 24, p. 93). 

Further complications in the distribution of the cords 
sometimes arise. Thus, in the tube of the Cherry, I find that 
the petaline cords assist in furnishing the calyx-limb with 
vascular cords, just as those corresponding to the arrested 
stamens of the Primrose enter the corolla of that plant. 
They either do not branch till they reach the angle between 
the sepals, or else from a point lower down. The small 
secondary branches are mainly directed outwards towards 
the margin, as represented in Fig. 29 ; s being sepaline, and 
p the petaline cords. 

In examining transverse sections of inferior ovaries, 
what one almost invariably observes is an inner epidermis, 
on some part or parts of which are placentas with ovules, 



an outer epidermis, and an intermediate ground tissue, 
apparently nearly uniform in character, from one epidermis 
to the other (as in Fig. 14, a to e, p. 68). A definite number 
of fibro-vascular cords jDeneti^ates this ground tissue. Theo- 
retically, if this structure consist of two parts, viz. the 
interior carpels and the exterior "tube," some line of 
demarcation might be expected to be traceable ; but in the 
majority of cases it would seem that, as neither the inner 
epidermis of the tube nor the outer one of the carpels are 
required, they are not developed at all ; and so the internal 
tissues of the two organs become confluent and uniform, and 
this accounts for the fact that the dorsal cords at least are 
simply embedded in this common tissue. Nevertheless, in 
some cases there actually is a certain differentiation in the 
tissue, as Van Tieghem has shown in the case of Alstrcemeria 
versicolor (Fig. 30), where a yellow band of cells marks the 

Fig. 29. — Receptacular tube and 
calyx-limb of Cherry. 

Fig. 30.— Alstrcevieria (after Van 

junction or congenital fusion of the two parts (indicated by 
the line in the figure). 

From the preceding descriptions, it will be seen, with 
regard to the sources of the cords belonging to the inner 
whorls, that they arise by division, radial or tangential as 
the case may be ; and then the secondary cords thus parted 
off are generally included within the tissue of the tube. 


These cords of tlie inner whorls may be given off at the 
terminal point of the pedicel; that is, at the base of the flower. 
In this case they may all run parallel from the base to the 
summit of the receptacular tube ; or they may branch at 
various heights within the tube itself, as in Prunus, described 
above ; or, lastly, they may not arise until the summit of the 
ovary is reached, when they pass off and enter their respective 
floral organs directly. These variations occur in both free 
receptacular tubes as well as when coherent to ovaries. 

As an example of the first case may be mentioned Alstron- 
meria versicolor; of the second, Galanthus nivalis, or Snowdrop; 
and of the third, Narcissus. In Alstroemeria, all the floral 
appendages have their cords distinct and independent, but 
invaginated by the tube of parenchyma throughout (Fig. 30). 
In the Snowdrop, the carpellary cords are distinct, but the 
perianth and androecium are inserted in the pedicel by a single 
verticil of cords, which becomes double higher up. Lastly, in 
Narcissus, all the parts of the flower are originally inserted 
in the pedicel by six cords, of which three give rise by 
successive tangential fission to a radial series composed of 
the dorsal cords of the carpels, the stamens opposite to the 
sepals, and the sepals themselves. Similarly, the other three 
form the petals together with the whorl of stamens opposite 
to them.* 

In Campanula, and to some extent in Lohelia, the cords 

* Ph. Van Tieghem, to whose researches I am indebted for the above, 
but which I have also paralleled in other cases, represents them neatly by 
the following formulas, wherein ( ) signifies vascular union, and [ ] the 
cellular union of the receptacular tube ; while (d) stands for the dorsal 
and (m) the marginal cords of the carpels. Stp signifies petaline and 
St. sepaline stamens. 

Alstr(smeria—13 S + 3 P + 3 St. + 3 Stp + 3 CJ. 

Galanthus—[S (S + St.) + 3 (P + Stp) + 3 C.]. 

Narcissus— IS (S + St. + d C.) + 3 (P + Stp) + 3 CJ. 


belonging to the petals are given off by radial chorisis from 
the sepaline, either quite from the base of the ovary or from 
about midway up the tube; they then diverge right or left at 
an acute angle, and, as soon as they have reached the summit 
of the ovary, pass up into the corolla.* As a rule, however, 
the petaline cords of flowers are quite distinct from the 
sepaline ; the six or ten, common to Monocotyledons and 
Dicotyledons respectively, forming the fibro-vascular cylinder 
in the pedicel. 

In all these and other cases the cords running up the 
receptacular tube proceed originally from the petiole, and 
are, so to say, even there intended for the appendages above. 
Normally they retain their axial character, in being arranged 
in a circle round the centre; abnormally an appendicular 
character can be revealed, by their becoming free and assum- 
ing a foliaceous aspect, as in Roses or Fuchsia, as mentioned 
above ; so that as long as the tube is normal, i.e. a cylinder 
of cortical parenchyma with cords, it is of the nature of axis, 
and can develop extra phyllomes and even buds ; but abnor- 
mally, the foliar nature, usually limited to the floral members 
at the summit, is extended to a greater distance lower down 
and the cords may now be converted into petioles, etc. 

Hence it appears undesirable to call it either a calyx tube 
or axial ; for these terms would seem to bind one to consider 
it permanently and in all cases as being either of one nature 
or the other. The term receptacular tube is therefore best, 
as it certainly " receives " or supports the whorls of the 
flowers ; and Teratology clearly shows that it can be either 
foliar (petiolar) or axial according to circumstances. 

* This reminds one of the way in which stipular appendages of 
Galium, etc., are supplied with cords — not by their intercalation into 
the common fibro-vascular cylinder of the stem, but — from a horizontal 
circular zone of fibres which connects the cords of the opposite leaves. 


Just as the two complete vascular cylinders of two separate 
floral peduncles can become fused into one oval cylinder when 
the latter are " fasciated," so, too, would it seem that the cords 
belonging to the separate parts of a floral whorl, where there 
is no receptacular tube, can form a single united cylinder, 
which one then designates as the receptacular tube. 

In the case of the inferior ovary, 1 would again emphasize 
the fact that the difficulty felt as to what is axial and what 
carpellary is entirely removed if the undifferentiated con- 
dition of the carpels be thoroughly understood. Indeed, 
whenever two organs are congenitally in union the epidermis 
of each is undeveloped, and the two mesophyls become one ; 
so that the dorsal cords of the carpels and those proper 
to the axis are alike plunged into a common tissue, which, 
regarded as one, is neither wholly axial nor wholly carpellary. 



The Form of the Perianth — General Observations. — It 
requires but a most cursory observation of flowers to notice 
how great is the variability in the forms of all their organs ; 
and the questions now before us are, how these morphological 
characters are correlated to the one process of pollination in 
order to secure the fertilisation of the flower, and how this 
infinite diversity of form has arisen. 

Most important differences in this respect follow from the 
fact of flowers being regular or irregular, and, when adapted 
to insects, according as the honey is easily accessible or not. 
Regular * flowers when borne singly are almost always 
terminal ;"!■ and when they are arranged in racemes, etc., 
they either stand out erect at the ends of their pedicels so as 
to be readily approached at any point of their circumference, 
as in the Wallflower, or else they are pendulous ; under which 
conditions, as a rule, no particular part is favoured by the 

* It is usual to speak of a flower as being regular or irregular ; but 
the term should be, strictly speaking, confined to one whorl at a time ; 
though when the corolla is irregular, the calyx and stamens are usually 
somewhat irregular as well. 

t The central and terminal flowers of many plants which elsewhere 
bear irregular flowers are often regular, as in Horse-chestnut, PeZarsromww, 
several of the 8cro;phularinecB, as Snapdragon, Linaria, Pentstemon, etc. 


insect more than another. It is only when the flower is 
situated laterally and projects horizontally, or approximately 
so, with its limb or border in a vertical plane, and, moreover, 
is more or less closely applied to the axis, that an insect is 
compelled to alight upon it on one side only, when approach- 
ing it directly from the fi'ont. It then throws all its weight 
upon the organs on the lower or anterior side of the flower, 
as is the case with the keel petals of papilionaceous flowers, 
with the lips of Labiates, etc. ; or else its weight is sustained 
by the stamens or style, or by both together, as in Epilobium 
angustifolium, Circcea, Veronica, Larkspur, and Monkshood; 
and whenever the stamens are declinate, as in Horse-chestnut, 
Dictamnus^ Echium, Amaryllis, etc. 

Flowers which have irregular corollas mostly show various 
degrees of " bilateral " form in their different whorls, and, 
have been called " zygomorphic." Such flowers, as a rule, do 
not receive the visits from so many differeut species of 
insects as regular flowers. These latter, not being charac- 
terized by the possession of any very definite contrivances for 
securing special insect agency, are accordingly visited by a 
much greater number and variety than those flowers which 
have become markedly adapted, and consequently restricted 
to particular visitors. 

It must not be forgotten, however, that regular flowers, 
if the tube leading to the honey be very contracted and more 
or less elongated, may become almost as much exclusive as 
very irregular ones ; for such flowers are mainly restricted to 

The following examples may suffice to illustrate these facts. 
Ranunculus acris, which is perfectly regular and with no 
specialized structure, is visited, according to Miiller, by more 
than sixty different species of insects ; whereas species of 
Aconitum and Delphinium, the two most highly differentiated 


and the only genera with irregular flowers of the same order, 
are adapted to, and mainly visited by the larger species of 
bees. Similarly of conspicuous and regular flowers of 
Rosacece, Primus communis has twenty-seven visitors ; Spircea 
Ulmaria, twenty-two ; JRuhus fruticosus, sixty-seven ; Fragaria 
vesca, twenty-five ; Cratcegus oxyaca7itJia, fifty-seven. On the 
other hand, of irregular flowers, Digitalis purpurea has only 
three useful visitors ; Linaria, nine or more species of bees, 
and Orchis mascula only eight. 

As an instance of a long-tubed regular flower, Lonicera 
ccerulea may be mentioned. It is adapted to humble-bees, by 
which it is chiefly visited. Similarly, the flower of the 
Honeysuckle, the lobes of which are scarcely if at all 
unequal, admits only a few lepidopterous insects which can 
reach the honey. So, too, Asperula taurina, which has a tube 
9 to 11 mm. long, is visited by nocturnal Lepidoptera. 

The Origin of Irregulaeitt. — With reference to the 
theoretical origin of irregular whorls, I assume that they 
have all descended from regular ones through external 
influences.* With regard to terminal, regular flowers the flow 
of sap ^is directed equally, radially, and in all directions on 
reaching the floral receptacle, and there is no inherent 
cause to make a terminal flower zygomorphic, or to induce 
one or more parts of any whorl to grow differently from the 
rest. Hence the primary cause of irregularity must come 
from without, and I regard this cause as issuing from the 
insect itself ; namely, the mechanical influence of its weight 
and pressures. To this external irritation the protoplasm of 
the cells responds, and gives rise to tissues which are thrown 
out to withstand the strains due to the extraneous pressures 

* The fibi'o-vascular cords of the pedicel are arranged at regular 
intervals, and are pei-fectly symmetrical around the medulla in irregular 
flowers, just as they are in the case of regular ones. 


of tlie insect, and so the flower prepares itself to maintain an 
equilibrium under the tensions imposed upon it, and irregu- 
larities are the result. Such, for example, occur in bilobed 
calyces, as of Furze and Salvia; in the many forms of " lips," 
or labella,* and enlarged anterior petals ; in dependent 
stamens, as of Aconite and Epilobiuvi angustifoUmn, or in the 
more usually declinate condition, as of Dictamnus, Amaryllis, 
etc. In these latter instances, in which the androecium bears 
the burden, the anterior petal is either, as a rule, unaffected, 
and shows no increase in size, or else there is a tendency to 
atrophy, so that it is reduced in size, as are the keel petals 
in Amlierstia. It is sometimes even wanting altogether, as 
in the Horse-chestnut.f 

* If the flower be resupinate, then it is the posterior organ which, 
now being in the front, has become enlarged ; as in V^ola and Orchis. 

f There has been more than one investigation into the causes of 
zygomorphism (as by Vochtung, Bei: Deutsch. Bot. GesselL, iii. (1885), 
p. 341 ; and Pringsheim's Jahrh. f. Wiss. Bot., xvii. (1886), p. 297 : also, 
by Dr. F. Noll, Arbeit Bot. Inst. Wiirzhurg, iii. (1887), p. 315). H. 
Vochtung distinguishes three different sets of causes as producing 
zygomorphism, viz. gravitation only ; gravitation acting on the consti. 
tution of the organs ; and the constitution of the organs alone. 

An objection to gravitation pure and simple is, that all flowers would 
be more or less subject to it, and become more or less zygomorphic 
accordingly. It does not account for the infinite diversity in the forms 
of zygomorphic oi'gans ; nor for the many correlations for insect 
fertilisation which exist between all parts of the flower. If to 
gravitation, however, we add the weight of the insect, which simply 
intensifies it, and couple with this the pressures exerted by the insect 
in various directions, then we have an adequate theory, which 
gravitation alone could not supply. When Vochtung speaks of ** consti- 
tution alone" as a cause, I presume he means hereditary effect. If so, 
I would quite agree with him, as zygomorphic flowers now grow to be 
such from purely hereditary influences. When, however, he would 
attribute the form of Epilohium angustijolium to geotropism, as the 
supposed cause of the lowermost petals bending upwards, and the 
stamens and style downwards (see Fig. 34, p. Ill), I do not see how 


Compensating processes tlius come into play, so that 
while some parts are enlarged others are diminished, the 
former always having to bear the strains, while the latter are 
free from them. Thus the lip of Lamium consists of one 
much-enlarged petal, which forms an excellent landing-place, 
but the two lateral petals, not being required, are atrophied 
to mere points. Similarly, while the two posterior petals 
enlarge to form the hood, presumably due to the backward 
thrust of the insect's head, the posterior stamen has vanished 
altogether. The gamosepalous calyx now furnishes its aid 
to support the slender tube of the corolla, not only by 
doubling its number of ribs, but by uniting them all together 
by means of a sclerenchymatous cylinder within the mesophyl. 

If the tube of the corolla be very strong and well able 
alone to support the insect, the adhesion of the filaments 
being also a powerful addition to its strength, then the calyx 
often remains polysepalous, as occurs in the Foxglove, 
Snapdragon, Fetunia, etc. 

If, instead of the anterior petal forming the landing- 
place, the tube of a gamopetalous corolla has enlarged so as 
to admit the ingress of an insect which partly or entirely 
crawls into it; then it is this tubular part which, more 
especially having to bear the strain upon it, bulges outwards, 
or becomes more or less inflated in form ; while the lip or 
amterior petal, not having to bear the entire burden, is not 
particularly enlarged, if it be at all. The Foxglove and 
Gloxinia, as well as Petunia to a slight extent, illustrate this 
adaptation in irregular flowers, while "campanulate" flowers 
afford examples amongst regular ones. 

gravitation can act in any other way than " downwards." Bat if one 
observes how a humble-bee suspends itself on the stamens while its 
body, so to say, thrusts the petals aside and upwards, we find a much 
more satisfactory interpretation in the theory I have proposed. 


If no more than tlie head of an insect enter the flower, 
then the corolla shapes itself to fit it. Thus Snowberry, 
Scrophularia, and Epipactis only admit the heads of wasps, 
which are the regular visitors of these plants. 

Other instances in which the limb is not much, if at all, 
enlarged occur in flowers especially adapted to Lepidoptera. 
Hovering, as they generally do, before the flowers, and in- 
serting their long proboscides while on the wing, there is no 
tendency to develop larger anterior petals, but the irritation 
affects the tube only, which thus elongates and contracts, 
resulting in little or no irregularity in the flowers, as in 
(Enothera biennis, in which the calyx tube has contracted, or 
in Honeysuckle, which has a tubular corolla. If bees or 
other insects visit the flower as well, then some degree of 
obliquity may result, as in Teucrium Scorodo7iia. 

Thus, then, may we get a rationale of the structure and 
form of floral organs, and their great diversity corresponds to 
a similar diversity in the insect world ; for the flower, if it 
be visited by many, will presumably take a form correspond- 
ing to the resultant of the forces brought to bear upon it ; 
if visited by few, it will shape itself in accordance with the 
requirements of its principal visitors ; and thus is it that 
while some easily accessible flowers receive many classes of 
insects, others are restricted to few, or even one ; and then 
the insect and the flower are so closely correlated as to almost 
impress upon one the idea that they were mutually created 
for each other ! 

The accompanying figures of Duvernoia adhatodoides * 
may illustrate my meaning. Looking at Fig. 31, a, alone (sup- 
posing we know nothing of insect visitors), one might ask. 
For what use is this great irregularity ? why and how has it 

* From a paper bj Mrs. Barber, Journ. Lin. Soc. Bot, vol. xi., 
p. 469. 



Fig. 31.— Duvernoia adhatodoides. 

come into existence ? And no answer is forthcoming. Now 
turning to Fig. 31, 6, we see one use at least. The weight of 
the bee must be very 
great ; and the curious 
shape of the lip, with 
its lateral ridges, is evi- 
dently not only an ex- 
cellent landing-place, 
but is so constructed 
as to bear that weight. 
Moreover, the two 
walls slope off, and are 
gripped by the legs of 
the bee, so that it evidently can secure an excellent purchase 
and can thus rifle the flower of its treasures at its ease. 

Irregular corollas are very numerous, but certain prin- 
ciples, traceable to insect action, govern their forms. In the 
first place, the side upon which the insect rests, or at least 
upon which its weight is thrown, is always enlarged, and 
mostly forms the landing-place. It is almost always the 
anterior petal; if, however, the pedicel or ovary has been 
too slender to support it, then it has sometimes become 
twisted, and the flower is said to be resupinate, so that the 
posterior petal becomes anterior in position, and is now the 
larger one, since it supplies the landing-place for insects, as 
in Orchis. Fumaria might be called semi-resupinate, as the 
corolla has only rotated through 90°. A slight modification 
occurs in the "Bee-orchis," Ophrys apifera, which is usually 
described as having a twisted ovary like a true Orchis ; but in 
this species it has scarcely if any twist at all; the flower, 
however, is bent over to the opposite side of the stem, so that 
while the posterior petal is still the labellum, the ovary has 
itself remained perfectly straight. 


The next point to notice is that when the anterior petal 
is enlarged, the posterior one or more often enlarges also, 
while a corresponding tendency to atrophy affects the lateral 
ones. This is seen in many species of Leguminosce, Scrophu- 
lai'inece, and Lahiatce, and in zygomorphic flowers generally. 
It occurs thus in the wing petals of many papilionaceous 
flowers, as is particularly well seen in Onobrychis. The 
immediate causes, I repeat, I would recognize in the weight 
of the insect in front, the local irritations behind, due to the 
thrust of the insect's head and probing for nectar, coupled 
with the absence of all strains upon the sides. In some 
papilionaceous flowers the wing petals form a landing-place, 
as in Indigofera and Phaseolus. Whenever this is the case, 
they too are enlarged, as the lateral ones are in Fig. 31, and 
undertake the duty impressed upon them. 

When, therefore, one finds as an invariable rule how the 
front petals enlarge when flowers are compacted and visited 
only from the front, and thus become irregular ; and as 
such often occur in orders where flowers are normally regular, 
as Iberis, Centaurea, Heracleum, etc. ; and, moreover, when 
the same phenomena appear in orders having no affinity 
between them, as in Labiates and Orchidecs; and are, indeed, 
to be found throughout the length and breadth of the floral 
world, one is justified in attributing such irregularities to a 
common cause, that being, according to my theory, the 
responsive power of protoplasm to the irritations from with- 
out, set up by insect and other agencies. 

Many other special cases might be described from the 
different orders of plants, but the above will suffice to illus- 
trate this principle of responsive action with resulting correla- 
tions to insect agency. I would here, however, call the reader's 
attention to the mechanical arrangement of forces as shown 
in Lamium and Echium, where it will be seen that the 


adhesions of tlie stamens to the corolla furnish the falcra, 
the cohesion of the petals into a tube affording a greatly 
increased power of resistance ; the weight of the insect on 
the labellum or declinate stamens is, of course, vertically 
downwards, and the line of the resultant, which the lip 
in Lamium and the stamens whenever declinate have to 
exert, passes through the point of meeting of the first 
two, and so sustains the insect while visiting the flower. 
Other and analogous instances will be described here- 

Good illustrations of the occurrence of great thickenings 
just where the strain will be most felt, may be seen in the 
slipper-shaped flowers of Calceolaria (Fig. 32), Coryanthes, 
and Cypripeditmi. Thus Calceolaria Pavonii 
possesses a thick ridge along the upper 
edges of the curved basal part, which 
carries the inflated end upon which the 
bee stands, and which it depresses to get 
the honey. In this species it may be 
noticed the anther-cells are separated (a), 
so that they can oscillate as they do in Fig. Z2.— calceolaria Pa- 

07* T n • T j-i- J • £ ^:^ z vonii (after Kerner). 

balvia. In Uypnpedium the edge is folded 
inwards, thus strengthening the same part; while in Cory- 
antlies the low^er portion is enormously enlarged, thus acting 
as a powerful spring which forces the anterior end of the 
labellum to be in close contact with the column. 

The Origin of Irregularity in the Axdrcecium. — As it 
is Tvith the perianth, so is it with the androecium : if the 
petals are regular the stamens are usually regular also ; but 
when irregularity occurs in the corolla the staminal whorl 
follow^s suit, and the position and form of the stamens are 
equally correlated to the effectual pollinatioa of the flower. 
Thus, as hypertrophy affects the anterior side of the 



flowers of Lahiatce, the anterior stamens are almost invariably 
the larger pair. On the other hand, atrophy has affected the 
posterior side of the stamina! whorl, causing the total logs of 
the fifth stamen, and, to some extent, a reduction in length 
of the next pair of filaments. 

When the weight of the insect is thrown upon the 
stamens, they either hang downwards, and the insect 
is suspended upon them, as in Epilohium angustifoUum, or 
else they become declinate and then the anterior petal, being 
relieved, does not enlarge, either remaining of the same size 
as the rest, or else diminishes, and may even vanish alto- 
gether. Thus Valuta, with its perfectly regular perianth 
and spreading stamens, may be compared with Amariilis, 
which has declinate stamens and a small anterior petal. 
The terminal flower of a " thyrse " of the Horse-chestnut, like 
the terminal flower of a "truss" of Pelargonium, is often 
regular ^ith spreading stamens, whereas the normal flowers 

have declinate stamens, and 
usually only four petals, the 
fifth or anterior one being 
altogether suppressed. 

In some flowers the sta- 
mens are dependent at first, 
but their anthers rise up when 
dehiscing, and so the fila- 
ments become declinate in 
the pollinating stage. This 
is the case with Delphinium, 
Epilohium angustifoUum,, and 
Didamnus (Fig. 33). In this 
flower the anterior petal is 
of much the same size as the others, but is often displaced 
(Fig. 33), and not immediately below the stamens, — this 

Fig. ^.—Didamnus (after Tleghem). 



lateral displacement of the anterior petal being not always 
carried out, as it is in tlie next flower to be described. 

In Epilohium angustifolium (Fig. 34) and Godetia, which, 
have no anterior petals, the bees cling to the dependent 
stamens, while the petals have become permanently displaced, 
the two lower being somewhat raised, so that the angular 
distances are not the same. In Azalea and Rhododendron 
there is no interior petal, but the posterior one is slightly 
enlarged, and this alone possesses extra colouring and the 
"path-finder." The stamens, being declinate, carry the 
insect without the aid of the corolla, so that the antero-lateral 

Fig. oL— Ejyilobium angustifolium. Fig. 35. — Veronica Chamcedrys (after Miiller). 

pair of petals, not sharing in the supjDort of the insect, are 
not enlarged at all. 

In Circcea and Vei'onica Chamcednjs (Fig. 35), the insect 
clings to the two stamens and style ; and the anterior petals 
are not enlarged, while in the latter flower it is, as usually 
the case, the smallest, the stamens of Veronica being attached 
to the lateral petals have to supply the fulcra for leverage, 
and consequently these have now become relatively hyper- 

In many flowers which have sub-declinate stamens, the 
latter lie in a more or less boat-shaped anterior petal, show- 
ing that the action of the insect has somewhat affected both 
the whorls together, as they have each some share in carry- 


ing the insect. Sucli is the case in the Ocimoidece of Lahiatce, 
in GolUnsia hicolor, the " Lemon-scented " Pelargonium, etc. 

Correlation of Growth. — I have only referred to the 
forms of flowers as grouped under the terms "regular" or 
" irregular," and alluded to a few instances ; for it is not my 
object in this work to merely give illustrations of various 
kinds, which are presumably well known to the reader, but 
to offer a rationale of the whole, without, however, attempt- 
ing to say how each individual shape has actually come into 
existence. To do this, it would be impossible in the present 
state of our knowledge of the history of flowers; my object 
being to suggest a probable cause, namely, the mechanical 
influence of insects, without excluding others which we 
cannot trace. Nutrition-, however, must be always borne in 
mind as an important one, hereditary influences as others — 
as, for example, in the restoration of an irregular flower to a 
condition of regularity, as occurs in Linaria, Lamiiim, Glox- 
inia, etc. The point, however, which I would specially 
emphasize is the correlation existing between the several 
parts of the organs, so that, regarded collectively, they all 
conspire to secure one and the same end, that being the 
pollination of the flower. Thus, as I have shown above, the 
calyx of Salvia has a form and structure correlated to the 
tube of the corolla ; the corolla has a form in strict adapta- 
tion to the weight and pressures of the insect which rests 
upon the lip. The stamens are, again, correlated to the pres- 
sures brought to bear upon them, and have grown in 
resiDonse, forming the remarkable lever-processes, which are 
also found in species of Calceolaria. Lastly, the style and 
stigma are correlated to the position of the anthers. Hyper- 
trophy in one direction has brought about atrophy in 
another, so that the two posterior stamens, are rudimentary, 
while the fifth has vanished altogether. 


Now, it might be argued, that when one organ changes 
its form others must do so in obedience to the " laws of cor- 
relation of growth," as IMr. Darwin showed to be the case 
with the feet and bills of pigeons. In plants, however, the 
connection between various parts, even in close proximity, is 
by no means so intimate as between different organs of the 
higher animals ; while the theory advanced here gives a 
common interpretation for the whole of the so-called correla- 
tions found in any flower. That one is justified in saying 
that correlated growths are much restricted in plants, is 
clear from the experience of horticulturists ; thus, while, 
e.g., the varieties of pease are infinite, they having been the 
object of selection alone, the flowers which produce them 
have virtually remained unchanged. 

A single coincidence has little or no scientific weight as 
indicating cause and effect. It is only when coincidences 
can be multiplied that they furnish a probability of a high 
order ; which, even if they do not admit of a verifiable ex- 
periment, still furnish a moral conviction, which, by the rules 
of philosophy, is equivalent to a demonstration. Now, this 
is exactly the case with irregular flowers. They always 
occur in similar positions ; they are always constracted so 
that the insect in adaptation to them can gain access to 
the honey in the easiest way ; their organs are so situated 
that the pollen should be transferred accurately to the 
stigma, etc. And when we find them distributed every- 
where throughout phanerogamous plants, the probability 
that the same or analogous causes have brought them about 
is of a very high order indeed 

Moreover, since we have abundant evidence of the re- 
sponsive power of protoplasm to build up tissues wherever 
they are required, I am not assuming an influence on the 
one hand without ample evidence of the probability of the 


responsive action on the other, coupled, of course, with here- 
ditary and other influences which fix the variation. Thus, 
then, as I believe, all flowers as we have them now, which 
are in perfect adaptation to insect agency, are the outcome 
of the resultant of all the forces, external and internal, which 
the insect has actually brought into play or stimulated into 
action by visiting them for their honey or pollen. 

The belief that such processes may have grown in 
response to mechanical irritations is supported by some 
interesting experiments made by Mr. O'Brien, of Harrow, 
who has kindly favoured me with the following remarks : 
" With reference to impressions conveyed by ' nervous ' force 
in Orchid flowers, whereby the expansion of the sepals and 
petals signifies to the reproductive organs that the time for 
fertilisation has arrived, I have observed that the periods of 
maturing and of decay may be either arrested or hastened in 
certain orchids by artificial means. With reference to arrest- 
ing decay, I took such flowers as Stanhopea and Corijanthes, 
which have large membranous sepals, and which, in the 
ordinary course of events, become reflexed soon after the 
opening of the flowers, and shortly afterwards wither. 
These are then followed by the other parts. By seizing the 
opportunity as soon as they expand, and by passing a thread 
round them, so as to keep them m the condition of the flower 
when just on the point of expansion, they may be kept good 
for a long time, the flowers evidently, as it were,^ not 
realizing the increased lapse of time, and being unaware that 
they had passed the period when they would have been ready 
for fertilisation. When so secured, a flower of Coryanthes 
speciosa on my table kept fresh three times as long as it 
would have done on the plant. The dripping of the water 
from the horns above the bucket is also arrested. Finally, 
on releasing the ligature, the broad wing-like sepals imme- 


diatelj became reflexed, and the water commenced to drip. 
Shortly afterwards the wings shrivelled up, and the flower 
decayed in the same manner as it would have done a week 
before if left to itself on the plant. 

" I will now give an example of deceiving a flower by 
artificial means, by making it believe that its fertilisation 
has been accomplished without its having taken place at all, 
Miltonia Russelliana carefully guards the approach to the 
column by closing the petals over it ; but on pushing these 
petals aside with a pencil, I always found that the labellum 
faded, and withdrew upwards very soon afterwards. The 
showy portion of the flower, evidently having had it con- 
veyed to it that its duty was performed, then followed suit. 
On carrying the deception still further to the reproductive 
organs, by placing small pieces of grit on the stigma, I 
found that the ovaries would swell in many cases, just as 
though the flower had been properly fertilised by pollen. 
This same result often takes place in Orchid flowers under 
cultivation, and seed-vessels are obtained of full size, but, of 
course, with no vitality in the grains within." 

As an analogous instance, I will add that it is the belief 
of M. 0. Beccari that ants are not only responsible for the 
remarkable growths in Myrmecodia and Hydnopliytum, etc., 
but that they have become indispensable for the healthy 
development of such plants. The investigations of M. Treub 
on Dischidia, the pitchers of which are frequented by ants, 
like the stipules of Acacia sphoerocephala, seem to justify one 
in concluding that genus also to be one of these so-called 
"Ant-plants " {Ann. du Jard. Bot. de Buitenzorg, iii., p. 13). 

Dr. Lundstrom also believes that the habit of producing 
" domatia " is now hereditary, without the actual presence of 
the insects (see Journ. Boy. Micr. Soc. 1888, p. 87.) 




Bilateral Symmetry. — A feature abundantly illustrated 
through the flowering world, in the construction of irregular 
flowers which are highly specialized for insect agency, and 
of which the LahiatcB and 8crophularinece, for example, fur- 
nish many instances, is the hypertrophy of the corolla in 
the direction of an antero-posterior plane, giving rise to a 
bilateral structure. 

On the one hand, the lips of various kinds, as also the 
keel, and often the wing petals too, where they help to sup- 
port the insects in papilionaceous flowers, are accounted for 
by the weight of the insects bringing about a responsive action 
in the protoplasm, thus determining a flow of nutriment to 
the parts demanding it, which now grow into the forms re- 
quired. On the other hand, the opposite or posterior side is 
often influenced as well, so that, as in Lamium, the lobes of 
the two posterior petals have grown into the enlarged hood. 
The cause of this I take to be the powerful tlirust which 
insects exert against the posterior side while their weight is 
expended on the anterior. If a humble-bee be watched, as 
represented in Fig. 31 (p. 107), it will be seen how eagerly 
and determinedly it forces its way into a corolla-tube if it 
expand upwards, as in Duvernoia or Lamium. All the pres- 
sure is exerted along the median plane, like an oblong wedge 


thrust into a circular tube. The corolla then " gives," as 
it were, and expands along the antero-posterior plane. The 
calyx follows suit, and often assumes a bilobed funnel-shaped 
tube as well ; while the lateral lobes of the corolla tend to 
atrophy, since they do not lie along- the line of the pressure 
due to the weight of the insect (see Fig. 40Z), p. 126.) 

If the floral organs be imagined to consist of some 
plastic, extensible, but not elastic substance, and be subjected 
to various pressures, strains, thrusts, etc., in imitation of the 
motions of insects, it is readily conceivable how the parts 
would yield, stretch, or bulge, and become fixed into shapes 
very closely resembling what has actually taken place in 
nature. In reality, of course, the ability to grow in response 
to the forces applied is to be substituted for the theoretical 
plasticity and extensibility of the imaginary material. 

Compensatory degenerations occur in various directions, 
as in the atrophy of the lateral petal-lobes of Lamium, the 
loss of the fifth posterior stamen, the reduction in length of 
the filaments of the posterior pair of stamens. In this latter 
respect Nepeta differs from other genera , but as we can 
readily conceive how all sorts of differences may and do exist 
in the direction and degree of the forces applied 
to flowers, some exceptional ones must have 
occurred in that genus which has favoured the 
growth of the posterior pair, so that they have 
become the longer ones ; for there is no rule 
without an exception. As another illustration, 
Teucrium may be taken. In this genus the 
*' hood " is entirely wanting ; but here, again, ^. „„ ^, 

•^ , ° ' 'to 5 Fig. 36.— Flower of 

the interpretation is that, no hypertrophy Teucriuvi (after 

^ , ' J f f J jiot. Mag., 1279). 

having been applied to them, the two petals 

of which it is composed have become reduced in size and 

"cleft," as shown in Fig. o6, of T. (Teucris) orientate. Bees, 


when visiting the flowers, hang downwards upon the corolla, 
as the lip and adjoining lobes are in one vertical plane, and 
give no thrust upon the posterior side. All weight, therefore, 
is thrown upon the front, just as it is on the stamens of 
Bpilohium angustifolium, described above. Their weight has 
consequently, so to say, " split " the hood in twain, and the 
stamens now stand erect in the cleft. 

The peculiar form of the corolla, with the whole of the 
limb dependent in a vertical direction, must throw the weight 
of the insect so much to the front, that the leverage will be 
at a considerable disadvantage — much more so than when 
the insect stands more directly over the tube of a corolla; 
which latter, in that case, is often strengthened by that of 
the calyx. To meet this difficulty the pedicel is curved over 
at the top, as may be readily seen in our common Wood- 
sage, and forms a spring, while hypertrophy has attacked the 
posterior side of the calyx, in that it now carries two extra 
7 marginal ribs, one on either side of the pos- 

^i m terior dorsal one, as shown in the accompany- 
d d ing diagram. This is exactly the reverse of 

^ ^ what occui'S in Salvia, and others which are 
much more strengthened on the anterior side, when the 
insect stands more directly over the centre of the flower. 

Additional aid is also gained by the tube of the corolla of 
Teucrium being resilient ; the anterior pair of stamens form 
two thick ridges, much aiding it in this respect ; the posterior 
pair, however, are, so to say, " sunk " into the tissue of the 
corolla as to be invisible in a transverse section. 

Transitional Forms. — We may sometimes, as it were, 
catch the formation of irregular and zygomorphic flowers in 
the process of formation; for it not infrequently happens 
that one genus will be irregular amongst its allied regular 
ones. Thus Verbascum and Petunia are transitional genera, 


and stand intermediate between Solanacece and Scrophula- 
rmece. The former genus lias a less zjgomorpliic corolla than 
many of the latter order, and also retains the fifth stamen 
m varying degrees of utility We might regard both these 
genera as Solanaceous, and on the road to acquiring zygo- 
morphism, but to which neither has yet fully attained. 

" The short-tube [of Verhascum nigrum'] widens out into 
a flat, five-lobed limb, which takes up an almost vertical 
position ; the inferior lobe is the longest, and the two 
superior are shorter than the lateral lobes, so that an insect 
settles most conveniently upon the inferior. The stamens 
project almost horizontally, but curve slightly upwards from 
the tube, and diverge slightly from one another ; they 
alternate w^ith the petals, and again the superior is the 
shortest, and the tAvo inferior longer than the lateral ones. 
. . , The style is shorter than the inferior stamens, and 
bent down slightly below them " 

From this description, taken from Miiller's work,* which, 
with slight modifications, would describe Petunia as well, 
the reader will see how these flowers fulfil the requirements 
of self-adaptation to insect agency ; and in every point of 
detail are they responding to the forces impinged upon 
them. The weight of the insect being well to the front, 
hypertrophy is commencing on the anterior side, while 
atrophy follows on the others, there being no special thrust 
as yet on the posterior side of the flowers 

There are many other genera and species which stand in 
intermediate positions between others, and it has always 
been a matter of doubt to systematists as to which they 
should be referred. The interpretation of their existence I 
take to be as here described, namely, that they are in an 
actual transitional state, brought about by insect agency, if 
* Fertilisation, etc., p. 429. 


fchej be flowers visited; or by fluctuating conditions of nutri- 
tion, if not ; and then, arrested in that state. 

A further remark on a significant point may be added on 
Petunia. In this flower, as in Verbascum, the limb of the 
corolla stands in a vertical plane, the anterior lobe is a 
trifle larger than the others, the five stamens have a slight 
tendency to be atrophied on the posterior side, while the 
stigma has become just so much displaced as to hinder self- 
fertilisation. This property is, however, by no means yet 
lost. Florists are aware of it, and find it necessary to self- 
fertilise, but not to cross, these flowers artificially to secure 
plenty of seed; Mr. Darwin corroborates this (Cross and Self, 
etc., p. 193). 

We have, then, here a case, but by no means an isolated 
one, in which the forms of the floral organs are undergoing 
a change, but the physiological characters of the essential 
organs have not yet been influenced by the external stimulus, 
so as to become more or less inert ujDon one another, as is 
sometimes the case in highly differentiated flowers. 

Indeed, it would seem to be a universal rule that morpho- 
logical changes are more readily acquired than physiological 
barrenness ; as by far the great majority of plants have 
retained their self -fertilising powers ; and, when they have 
lost it, it is easily and rapidly reacquired when the necessary 
conditions are supplied. 

Echium is another instance of aloQOst a single genus 
amongst others of the same order characterized by great and 
persistent regularity. Wiododendron and Azalea may be 
compared with other genera of Ericacece, and the reader will 
readily suggest others. 

Sometimes the irregularity is confined to the stamens or 
style, or both, which may have a tendency to become decli- 
nate, as in Calluna, in some Liliaceous and Amaryllidaceous 



plants, as Narcissus Corhularia. In Anagallis arvensis and 
Lyclum barharum there is nothing but an obliquity in the 
style observable. 

In all the flowers which tend to show irregularities the 
rule is that the corolla-limb stands in a vertical plane, so 
that the flowers are visited from the front. This I take, as 
mentioned above,, to be generally a primary necessity for 
bringing about irregularities of all kinds. There are some 
campanulate and pendulous flowers where irregularity occurs 
in the lengths of the filaments or the size of the anthers. 
Thus, I have observed great fluctuations in the stamens of 
Narcissus cernuus : some of these I have illustrated in Fig. 37. 
I noticed that a peduncle always 
bore the same form in every 
flower of its umbel. There 
were mostly three flow^ers in 
each, as of a, h, and d; one 
specimen of a and one of e had 
only a single flower ; and one 
of c had two flowers. In a, 6, 
c, d, the three short stamens, 
as well as the three long ones, 
were all of the same height, respectively ; but in e one of the 
shorter set was taller than the rest. 

Similar fluctuations are not at all uncommon in cultivated 
heterostyled plants, as Primroses ; as will be alluded to again 
in discussing the conditions of heterostylism. 

In Fritillaria Meleagris, though no irregularity occurs in 
the perianth leaves, it often appears in the androecium, and 
is more especially observable in the lengths of the anthers. 
This would seem, therefore, to be another instance of 
incipient change. 

Calluna vulgaris is likewise just^ commencing to be 


Fig. 37. — Xarcissus cernuus. 


irregular. The flowers are almost horizontal, closely com- 
pacted against the axis, and consequently not readily visited 
on any side except from the front. The style and stamens 
curve upwards, so that " the smaller bees and flies thrust 
the head or proboscis from the front into the flower, and the 
upward curvature of the style and stamens causes the insect 
to enter by the lower half of the flower, and so to get dusted 
with pollen from above." * 

Miiller also notices, about this flower, that " the style, 
which even in the bud overtops the stamens, grows very 
markedly after the flower opens, as the flower itself does. 
As a rule, it attains its full length only after the anthers 
have completely shed their pollen, at which time also the 
four-lobed stigma reaches its full development." 

He gives five figures of Saxifraga Seguieri to show the 
progressive stages of development. In the first or female 
(protogynous) condition the stigmas only are mature, the 
anthers, petals, and sepals being far from having attained 
their full size. It is not until half the anthers have shed 
their pollen, and the others ready to do so, that the flower 
attains its complete dimensions f 

I refer to these facts, which are equally applicable to 
many other flowers, to show that growth normally continues 
after insects have commenced to visit flowers ; so that there is 
plenty of opportunity for the petals, stamens, etc., to respond 
to the insect's action before reaching maturity. 

Dr. F. Noll has investigated the various movements of 
zygomorphic flowers during growth, resulting in the external 
position of the flower; and he finds that the excess of 
weight on one side is, when necessary, counterbalanced by 
active tensions (see Jl. B. Mic. Soc, 1887, p. 612 and reffs.). 

* Fertilisation, etc., p. 379. f Ihid, p. 244. 



Yegetative Organs. — In explaining the origin of irregular 
flowers bj insect agency, it will not be amiss to fortify the 
theory by describing other instances apart from flowers, and 
to add further results which I believe to accrue from the 
persistent action of insects on the one hand, and a ready 
response on the part of the organ on the other. 

Researches into the anatomy of stems have proved the 
existence of this responsive power. Thus, a tree will develop 
wood in a particular direction if it be compelled to meet 
special strains imposed upon it ; for Andrew Knight found 
that when trees were allowed freedom in one direction only, 
and were thus made to oscillate in definite directions, either 
east and west or north and south, the stem became elliptical 
in section, the long axis corresponding to the direction of 
oscillation. Mr, Herbert Spencer has also described how 
Cactuses, if submitted to particular strains, develop wood to 
meet them. 

The various kinds of the supporting tissues of pedicels, 
such as collenchyma, sclerenchyma, the so-called liber-fibres 
as well as true woody fibre, are all so many contrivances of 
the stems to support the weight of the flowers and fruits, and 
to overcome gravity. So, again, in the case of apples and 
pears, if they hang vertically downwards they grow as 


symmetrically round the insertion of the stalk as an orange ; 
but if the pedicel projects obliquely from the branch, they 
then thicken along the upper side, forming a sort of buttress 
running^ down into the stalk, which also itself tends to 
thicken. This enlargement, which gives the peculiar " lop- 
sidedness " to several kinds of pears especially, and in a 
lesser degree to some sorts of apples, is simply due to the 
fact that the force required to counteract the resultant of 
the two forces, gravity and tension — which act vertically 
downwards and along the stalk, respectively — must be 
increased in proportion as the direction of the stalk ap- 
proaches the horizontal one. The accompanying diagram 
(Fig. 38) represents the basal end of a Dr. Jules Guyot pear 
and in the position in which it -hangs 
-^ upon the tree. The letter it; (weight) 
is in the line of gravity, t (tension) 
acts along the stalk, w:hile r coun- 
teracts the resultant, which tends to 
tear the pear from the stalk at the 
upper side. This strain must be 
J, met, and the increased thickness 

Fig. 38.— Diagram of the end of a along this Upper sidc enables the pear 

Dr. Jules Guyot pear. ^^ ^^^^.^^ .^^ ^^^ ^^^^ prevents the 

fruit, especially if it be a large and heavy kind, from being 
wrenched from the stalk. 

A somewhat similar development often occurs with plums 
and lemons ; only, as there is no receptacular tube in either 
case, the weight of the fruit causes them to produce a thick 
fold in the carpel on the under side, together wath some 
degree of hypertrophy on the upper, where the tension occurs. 

It is not uninteresting to notice how branches of trees 
similarly sustain the strain produced by their own weight. 
This is done by growing at an acute angle (originally caused 



by arising in the axil of a horizontally inserted leaf), much 
more often than in a strictly horizontal direction. The 
branch, after growing for a short distance upwards, generally 
bends downwards, assuming just the same curvature as of 
declinate stamens which have to support the weight of 

If the vertical line in the adjoining diagram (Fig. 39) 
represent the trunk, and the curved 
line a branch, the insertion at /sup- 
plies the'fulcrum, w is the weight of /I 
the branch, and acts in a vertical I \vA 
line, p is the power required to 
counteract the resultant of these two 
forces. ^ 

Whpn fhp Tinncrli Lrpalc«? pU>,p-p Fig. 39.— Diagram of a tree and 

vv nen rne oougn oreaKs, eitnei ij^anch, illustrating the distri- 
through an additional weight of snow ^'^^^^^ offerees. 
or by its own weight on decay, it snaps off at the point p, 
i.e. the place where the force acts, as it can no longer over- 
come the resultant of / and w. 

Reproductive Organs. — Applying these principles to 
floral structures, we have already seen in how many Avays 
the strain to which parts of flowers are subjected, through 
the weights and pressures of insects, are met and overcome. 

In a large number of instances the organ becomes curved, 
and assumes the character of a spring, yielding on pressure, 
but recovering its position when pressure is removed. It is 
often so with the claws of the petals of papilionaceous 
flowers, the stamens of Dicenira, CorydaUs, and Veronica 
Chamcedrys. Similar structures are seen in many styles, as 
those of Pansy (Fig. 54), and in genera of Polygalacece. 

All declinate stamens partake of it to a more or less 
degree. The distribution of the forces brought into play to 
support the insect is exactly the same as when a bough 



has to support its own weight, as will be easily understood 
from what has been described, and by referring to the 
diagi'am (Fig. 40a). 

If the tissue does not remain firm 
under pressure, then the lever-action 
of a spring may fail to be secured, 
and the organ will oscillate freely, 
as on a pivot. This 1 take to be 
another result of a constant, but of 
course unconscious, effort of the insect 
Fig. 40a.-D!a,crram of deciinate ^^ P^^^ ^^^ Organ in a Certain direc- 
bu\Toa"Jf force?^'"^ ^^^ '^''*"' ^^^^' ^^ '^^ ^^^^ ^^^* anthers become 
versatile, and oscillate, and may 
become even inverted in position, when pollination is being 
effected by insects. Consequently anthers normally introrse 
can be made to assume a pseudo-extrorse position. This 
happens with some Cruciferce as Cardamine pratensis, Tulips, 
etc. A similar cause I would attribute to the formation 
of the oscillating anthers of Salvia, and of the species of 
Calceolaria^ as G. JPavonii, which form the 
section Aposecos of that genus, as shown m 
Fig. 32, a, p. 109. 

As an example of an entire flower illus- 
trating the distribution of forces, the accom- 
panying figure of Lamium album (Fig. 406) 
will explain how the forms of the calyx and 
corolla are adjusted to bear the weight of 
the insect. The bee alights on the lip and 
then partially crawls into the expanded 
mouth of the corolla, so that its weight 
now lies in the direction of w. The fulcrum will be at /, 
and the I'esultant of these is in the opposite direction to r. 
This is where the strain will be felt ; so that it is just at this 

Fig. AOb. — Lamium at 
bum, ghowing distri 
button of forces. 


point where the backward curvature talces place which gives 
strength to the corolla-tube. This latter is also greatly 
supported by the tube of the calyx, which, as stated, has a 
curiously thickened cylinder within the mesophyl. 

Finally, if we may admit the existence of this adaptability 
to strains. and other external forces, and that the various 
structures of fio^vers will grow in response to them and 
develop themselves accordingly, we have a clue to the 
interpretation of every one of the most diverse forms which 
may be found in flowers adapted to insect agency. 

Similarly, with regard to several classes of cell structure 
which are now recognized as having a supportive function, such 
as collenchyma, sclerenchyma, w^ood fibres, etc., I would con- 
tend that such are not formed originally and anteriorly to the 
requirements of the plant; but that strains have beenresponded 
to, and the tissues formed accordingly. Then, subsequently, 
hereditary influences have come into play, so that noiv they may 
appear even before there is any actual necessity for them. 

I find that M. J. Baranetzki's observations * on the thick- 
ening of cell-walls tend to corroborate this view ; for he, too, 
has arrived at the conclusion that the secondary formations 
on the inierioi' of the cell-ivalls are always in adaptation to 
protect the cell- wall against the pressures exercised upon it. 

In alluding to the above instances of levers and mecha- 
nical powers in plants, one mentally recalls how abundant 
they are in the distribution of the bones and muscles in 
vertebrates. These latter are, of course, situated only and 
exactly where they are required. I cannot help thinking, 
therefore, that the old view was fundamentally correct; that 
such, have been gradually brought into existence by the 
efforts to meet the strains put upon them. If this be true, 
then one and the same law has prevailed in the evolution of 
organs in both the animal and vegetable kingdoms. 
* Aim. des Sci. Nat (Bot), iv. (1886) p. 135. 




Reversions to Regularity. — Dr. Masters observes that "in 
cultivated Pelargoniums, the central flower of the umbel or 
* truss ' frequently retains its regularity of proportion, so 
as closely to approximate to the normal condition in the 
allied genus Geranium ; this resemblance is rendered gi^eater 
by the fact that, under such circumstances, the patches of 
darker colour characteristic of the ordinary flower are com- 
pletely wanting, the flower being as uniform in colour as in 
shape. Even the nectaiy, which is adherent to* the upper 
surface of the pedicel in the normal flower, disappears, some- 
times completely, at other times partially. The direction of 
the stamens and style, and even that of the whole flower, 
becomes altered from the inclined to the vertical position. 
In addition to these changes, which are those most commonly 
met with, the number of the parts of the flower is sometimes 
augmented, and a tendency to pass from the verticillate to 
the spiral arrangement manifested." * 

All the differentiations in an ordinary xateral blossom of 
Pelargonium brought about by insect agency are, in the above 
instances, reversed in consequence of the terminal position 
of the flower. A more complete illustration of the effect of 
manner of growth and the distribution of nutrition could not 

* Teratology, p. 221. 


well be given, showing how all the features of irregularity 
acquired by the ordinary form must have been induced or 
impressed upon the flower when growing laterally and easily 
visited, but that they are readily lost as soon as the sap 
can be distributed radially and so cause the parts to grow 
symmetrically round the now vertical axis. 

Besides the occasional appearance of one or more terminal 
and regular flowers among a truss of in^egular ones, it is the 
object of florists to induce all the blossoms of many iiTcgular 
flowers to become regular. Thus cultivated Pelargoniums, 
Gloxinias, Azaleas, Pansies, etc., which are normally irre- 
gular, tend to become regular under cultivation, and lose their 
characteristic features. 

In all these cases I am inclined to recognize negative 
evidence in favour of the theory advanced; in that, presuming 
the characteristic irregularities to have been brought about 
by the agency of insects and through the crossing of distinct 
flowers by these creatures, and that the irregularities have 
arisen under the various pressures, etc. ; then, under cultiva- 
tion, though they may be repeatedly crossed by man — the 
process, however, not being effected in the same way as by 
insects, and consequently the causes of irregularity being 
wanting — the flowers now revert to their ancestral forms; 
while ample supplies of nutriment doubtless play an important 
part in the process. 

Moreover, though any irregular flower may become regu- 
lar, it is a significant fact that normally regular flowers are 
never known to suddenly assume any definite irregular form. 

That the change from ii^egularity to regularity is an 
acquired constitutional affection is seen in the fact that, 
when the flowers of a drooping Gloxinia are fertilised with 
their own pollen, a large number of the seedlings will bear 
the erect regular form of flower. 


In the preceding cases tlie regularity occunnng in 
normally ii'regular flowers is dae to the non-development 
or arrest of the usually characteristic features which give 
rise to the irregularity ; so that the resulting form is a 
reversion to, or a restoration of, the ancestral conditions of 
the flower which is assumed to have been perfectly regular. 

As insects, by their mechanical actions, are here believed 
to have brought about irregularities in flowers ; so, con- 
versely, regularity can be reacquired through their agency 
in another way. Clerodendron is a plant in the corollas of 
which certain members of the family Tingidce take up their 
abode as pup«. The irritation induced by their presence 
brings about a hypertrophy of the corolla, which now 
assumes a regular form, while the filaments and style are 
likewise affected, becoming much thicker than in the normal, 
irregular flower. 

Reversions to regularity may, therefore, I think, be safely 
referred to nutrition as the immediate agent, though such 
extra flow of nutriment may be brought about by diverse 

"Peloria." — Regularity may, however, arise in another 
way, by the members of the whorl or whorls normally 
irref^ular being all exactly alike. Instead of there being 
any arrest, there is here an excess of development. Thus, 
if, instead of the anterior petal of Linaria being the only one 
provided with a spur, all the petals become spurred, then the 
corolla will become regular; but there is no other tendency 
to revert to the ancestral form. This variety constitutes the 
form called " Peloria " by Linnaeus. 

There are, then, two factors, which appear either singly 
or together, in this process of change. First, a terminal 
position, as this tends to produce regularity in consequence 
of an eqaable flow of sap in all directions : just as this also 


determines the persistent regularity of all flowers wliicli are 
normally so situated and are visited from all directions. 
It will be often found that when Snapdragons have pelo- 
rian blossoms they are in three-flowered cymes as in Cal- 
ceolarias, instead of a raceme, of which the central one 
is regular, while the lateral flowers are irregular. Secondly, 
whether terminal or not, the influence which first brought 
about the change in the anterior part of the flower spreads 
to and effects all the rest. This statement, of course, only 
expresses what one sees, without explaining the process ; 
but the fact that the energy peculiar to the formation of one 
organ can affect others is so common, that we may recognize 
the process as a principle of growth; just as stamens may 
become petaloid, on the one hand, or pistiloid on the other ; 
showing that "petaline energy" can affect the androecium 
in the first case, and " pistiline energy " in the latter. 

That the true pelorian form is correlated to vegetative 
energy is seen in the fact that such a flower obviously requires 
more material than a normal one, and that petalody of the 
stamens frequently accompanies the modification. Moreover, 
although of course usually sterile under such circumstances, 
yet pelorian Linarias have been reproduced when the seeds 
were sown in a rich soil. Mr. Darwin also raised sixteen 
seedling plants of a pelorian variety of Antirrhinum artificially 
fertilised by its own pollen, all of which were as perfectly 
pelorian as the parent plant. 

That peloria is due to hypertrophy is also seen in the fact 
that it always arises by multiplication of the normally enlarged 
organ. Thus, in Linaria and AntirrJihium all the petals are 
spurred or pouched ; in pelorian Larkspurs and Aconites it 
is the spurred and hooded sepal which is repeated ; and in 
papilionaceous flowers it is the standard which is multiplied 
five times, etc. An abnormal increase in the number of petals 


and stamens often occurs in pelorian Pelargoniums, Horse- 
chestnut, etc. 

If pelorian forms were equally constant as the one-spurred 
condition, botanists would undoubtedly have recognized them 
as species, or perhaps genera, as it is the comparatively 
slight difference in the length of the spur upon which they 
separate Linaria from AntirrMnum. Similarly Corydalis has 
normally but one spar and one nectary. It, however, bears 
occasionally two spurs and has two nectaries, as in Bicentra. 

" Peloria, then," as Dr. Masters observes,* "is especially 
interesting, physiologically as well as morphologically. It is 
also of value in a systematic point of view, as showing how 
closely the deviations from the ordinary form of one plant 
represent the ordinary conditions of another ; thus the peloric 
' sleeve-like ' form of Calceolaria resembles the flowers of 
Fabiana, and De Candolle, comparing the peloric flowers of. 
the ScTophulariacece with those [the normal ones] of SolanacecB, 
concluded that the former natural order was only an habituRl 
alteration from the type of the latter. Peloric flowers of 
Papilionacece in this way are tindistinguishable from those 
of Bosacece. In like manner we may trace an analogy between 
the normal one-spuiTed Delphinium and the five-spurred 
Aquilegia, an analogy strengthened by such a case as that of 
the five-sparred flower of Delphinium.'' 

* Teratology, p. 236. 



Epidermal Trichomes, etc. — While all conspicuous flowers 
inyite insects of some sort or another to visit them, which, by 
so doing, pollinate their stigmas, it is an important thing to 
be able to exclude those which w^ould rifle the flower of its 
treasures and yet not transfer the pollen from one flower to 
another. Dr. Kerner, in his interesting work entitled Floivers 
and their Unbidden Guests, has described and figured a large 
number of instances of the forms of floivers in which he 
detects various processes, some of which produce sticky 
secretions, others occurring as hairy " wheels " and " tangles " 
of wool, etc. ; all of which tend to stop the ingress of ants 
and other small insects, and thus prevent them from getting 
at the honey. The question at once arises. How have these 
processes been caused ? Without attempting to account 
for all, the theory I offer will, I maintain, be answerable 
for a good many, especially for several cases of secretive 
processes and for the hairy obstructions. All these I would 
suggest as the immediate results of the irritations set up by 
insects ; so that, as a consequence, they occur just and only 
where they are wanted ; so that, while they form no hindrance 
to the larger and stronger insects which have presumably 
caused them to be developed, they, however, may effectually 
prevent the smaller ones from ertering. 



In many cases tbe capability of the flower to restrict 
itself to its proper visitors, and at the same time to exclude 
the wrong ones, is a common result of the differentiations 
which have taken place. Thus, an elongated tube, as in 
Evening Primrose, and in some species of Narcissus, etc., is 
a direct result of and adaptation to the long proboscides of 
Lepidoptera, and in proportion as the tube is elongated so 
does it prevent the ingress of short-tongued insects, or of 
those with short proboscides. 

Apart, however, from such and other general results of 
adaptations, whereby flowers have become, for example, 
irregular, and consequently their insect visitors are more 
and more restricted in number, there are innumerable out- 
growths of various kinds which act as special obstructions 
to the entry of small insects which would not be able to 
pollinate the flower. Thus, while many regular flowers, such 
as Gentians, have developed horizontal hairs all round the 
entrance to the tube of the corolla, Honeysuckle and Veronica 
Chamcedrys, which are irregular and approached from one side 
only, have developed them in the anterior side alone. In 

Ainiaryllis belladonna Kerner describes 
and figures (Fig. 41) a one-sided flap 
growing out of the perianth, and so 
folded as to fui^nish a very small orifice 
for the entrance of a proboscis. There 
is no such gi'owth on the anterior side, 
but only on that one, the posterior, 
which is probed by an insect. 

In Gentiana Bavarica there are 

Fig. 41.— Base of flower of ^ma- .,,,., i ii i f 

ryiiis showing houey-protector tooth-like processcs at the entrance 01 

er verner;. ^^^ tube, which remind one of the 

appendages to the corolla of some of the Silenece. Monotropa 

glabra and Daphrte Blagayana agree in having a large circular 


stigma nearly blocking up tlie tube ; and wliile in the former 
the irritation set up bj tlie proboscis of an insect has 
(presumably) given rise to a glutinous secretion, in the latter 
it has caused a development of hair.* 

Did we but know what the insects were, and how they 
have poised themselves upon the flower, and in what way 
their proboscides and tongues have irritated the different 
parts, one might be able to describe more accurately the 
whole process ; but that such has been the cause and effect, 
as above described, seems to me to be too probable a theory 
to be hastily discarded in the absence of a better one. 

It is one of those arguments of deduction that escape 
the opportunity of verification, and can only rest for support 
upon the number of coincidences which can be found, and 
which collectively furnish a probability of a high order. 

When, then, we find that these processes always occur 
just where we know the heads, legs, bodies, and proboscides 
or tongues of insects habitually are placed and irritate the 
flower, we are justified in recognizing, not only a coincidence, 
but a cause and effect, though we may not be able to trace 
the action in each individual case. Thus, it may be asked, 

* The remarkable fact of Heliotrope being the solitary exception 
out of the order Apocynacece, with the stigma forming a circular rim 
heloiv the summit, may meet with its interpretation from a like cause. 
The corolla is so folded round the style that it leaves no space between 
it and the latter. Hence it may, perhaps, have been due to a similar 
" rubbing," that has transferred the stigmatic surface from the now 
abandoned apex to a lower level, just where the style-arms ought to 
begin to diverge. The papillae, too, differ from the ordinary form in 
being pointed like fine hairs. The relative diffex'ences in the distribution 
of the papillae on the style-arms of the Compositee, I would also suggest 
as having been brought abont by different insects which irritate them 
in various ways. So, too, the diverging stigmas of insect-fertilised 
cruciferous flowers may be compared with the small globular form of 
self-fertilising species of the Cruciferce. 


Why are the three anterior petals of Tropceolum fringed, but 
the two posterior, which stand a long way behind, not so ? 
Why are hairs produced on the anterior side of a Honey- 
suckle and Veronica, but all round the mouth of the regular 
Gentiana ? And many other questions of a like sort might 
be raised. If we watch the habits of insects with their 
tongues, we may easily see how they irritate the various parts 
by licking them, not solely where the honey is secreted, but 
the filaments, etc. Thus Miiller often watched Rhingia 
rostrata licking the staminal hairs of Verhascum pJiceniceum, 
and in many cases the hairs on the filaments offer a foothold 
to the insects while visiting the flowers, as in species of 
Mullein ; such hairs, if my theory be true, being the actual 
result of the insects clutching the filaments or rubbing them 
with their claws. In Centaurea, the epidermal cells of the 
filaments hasp produced projectirg processes just where the 
proboscis rubs against them when searching for honey in 
the little cup (see Fig 11, p 60), from the middle of which 
the style issues, as shown by the direction of the arrow. 

These filaments also exhibit their extreme irritability by 
contracting, and so assisting in the " piston action " by 
dragging the anther-cylinder downwards over the style. 

While recognizing the coincidence between the localiza- 
tion of outgrowths, enations, trichomes, etc., and the position 
of the parts of insects in contact with flowers when searching 
for honey, one must not forget that a great number occur 
where such contacts do not take place. Hence we must look 
for other possible causes for their origin as well. One of the 
commonest forms of trichomes is glandular hairs, and, as Dr. 
Kerner has pointed out, when they occur on sepals, pedicels, 
etc., they form admirable barriers to the approach of ants 
and other creeping insects, which might rifle the flower and 
yet not fertilise it. We must be on our guard, however, in 


asserting that nature lias produced them in order to keep ants 
off ; for that line of reasoning is pretty sure to land us in 
faulty teleological methods. Wkat causes them is not at 
present known in all cases ; though we may perceive that 
certain conditions, as growth in water, can bring about their 
disappearance, as Dr. Kerner remarked in the case of 
Polygonum anipMbium, which only has them when growling 
on land. 

If, however, we ask, for example, Avhy the Sweet-briar has 
them all over it, and why the Dog-rose has none, I do not 
know how to reply to the question as yet. We may notice 
certain coincidences, that hairy herbaceous plants are com- 
moner in dry situations and smooth ones in watery ; just as 
root-hairs occur in a loose sandy soil and their absence is 
noticeable in a heavy one; but we do not know how these 
di-fferent media actually bring about these changes, though 
wx' may feel assured that it is solely due to the environment. 

If we, thus, look elsewhere than in flowers for any 
analogous processes they are by no means w-anting. For 
example, it is simply the mechanical irritation brought about 
by contact with a foreign 
body, probably aided by 
moisture and a lessened de- 
gree of light, that causes 
the epidermal cells of the 
aerial roots of the Ivy and 
Orchids (Fig. 42) to elon- 
gate into adhesive or clasp- 
ing hairs, so as to grasp the ^ig. 42.-Adhesive epidermal ceUs of roots of 

body for support. This is jancztwsS/''"^' ' ''' ^"''^''"■'■''"^*'' ^''^*^'' 
only a form of the ordinary 

root-hairs which are immediately developed w^hen the tip is in 
contact with a moist soil, and each hair grips and glues itself 


to the particles of soil.* Cliatin noticed the production of 
hairs when the roots came in contact with any obstacle ; t 
but Dr. M. T. Masters observes that the obstacle alone in 
their case is insufficient without moisture, for he found that 
the roots of Mustard-seed could penetrate a stiff clay, but did 
not develop any root-hairs until they came in contact with 
the sides of the pot — " Wherever there was a thin film of 
water investing a stone or the sides of a porous flower-pot 
or a plate of glass, there the root-hairs abounded." 

Besides a nutrient or moist medium, actual growth in 
water may enormously increase the length and quantity of 
root-hairs ; as may be seen in the dependent roots of floating 
plants of Hydrocharis, etc. ; or in the hypertrophied con- 
ditions of the roots of grasses when growing in water. 

That epidermal trichomes may be due to the irritation of 
insects is clearly seen by their appearance within the cavities 
of certain galls. J In the case, for example, of a very com- 
mon one on willows, the leaf bulges out below and forms 
a sort of bag, open or closed above. The tissues become 
hypertrophied though the epidermis and palisade cells are 
still recognizable lining the cavity. The leaf has scattered 
hairs on both sides ; but within the cavity much larger hairs, 
rich with protoplasmic or other matters, project from all 
sides into the interior. Some are straight, others curved, 
club-shaped, or wdth irregularly swollen ends, not unlike the 
forms produced on climbing roots by contact with a foreign 
body. Again, the crimson " spangles," so common on the 
underside of Oak-leaves, are covered with stellate clusters 

» Sachs' Phys. of PL (Eng. ed.), 1887, fii?. 12, p. 19. 

t Mem. Soc. Nat. Sci., Cherbourg, 1856, p. 5; referred to by Dr. 
M. T. Masters in Notes on Root-hairs, etc,,Jonm. Roy. Hort. Soc, vol. v., 
p. 174. 

X Caused by species of Nematus. 


of liairs. Similarly, those of Cecidomyia Ulmarice on Spircea 
TJlmaria are hairy outside, and papillose within ; while similar 
ones of a Pliytoptus on the Sycamore are lined with long 
blunt-ended hairs, and are clothed without by others, long 
and pointed. In all these cases the galls, as well as the 
hairs, are the product of irritation set up by the presence of 
the egg deposited by the insect.* 

As another very common instance of the presence of 
epidermal papillae and hairs, may be mentioned their occur- 
rence in the stylar and ovarian cavities. The former, and 
the placentas especially, may be clothed witli delicate hairs 
exactly resembling root-hairs. Such may be well seen in 
the Poplar, Tamus, Richardia JEthiojnca, etc.; and since M. 
Guignard f has discovered that the mechanical and physio- 
logical irritation of the pollen-tubes is required to cause 
their development on the walls of the ovary in Vanilla, 
between the longitudinal bands of conducting tissue, it is, 
I think, a by no means improbable theory that the tufts of 
hairs over the nectaries, " tangles," " wheels," etc., on the 
filaments or corolla- tubes, have been actually caused by the 
irritation of insects, since they occur just where such irrita- 
tions are made. 

One use of certain outgrowths has been regarded as 
intended to protect the honey from rain. Why, however, 
some flowers should be so favoured while many others, as 
of the Umhelliferm, have no protection at all, is not stated. 
The interpretation I have here offered will, of course, apply 
to all such growths, whenever they may really keep off rain 
or " unwelcome guests." 

* Krasan has lately discnsFed the formation of the woolliness of 
galls, etc., Oesferr. Bot. Zeitschr., xxxvii. (1887), pp. 7, 47, 93, seqq. 

t Sur la Pollinisation et ses Effets chez les Orchid^eSy par M. L. 
Guignard, Ann. des Sci. Nat., torn, iv., 1886, p. 202. 





Position of Nectaries.* — These honej-secreting organs 
seem capable of being formed anywhere. Of course they 
are mainly to be found in flowers, but many plants bear 
them elsewhere. Thus, some ferns have them on the rachis ; 
the common laurel, as also the almond and peach, have two 
at the base of the petiole ; beans and vetches, as well as 
species of Impatiens, have them on the stipules, as shown in 
Fig. 43. Bees may be often seen as busy about the young 

shoots of laurel as 
if they were visiting 
flowers. Acacia 
spliceroceioliala has a 
large one, on the 
upper side of the 
petiole, which sup- 
plies those ants with 
food which take up their abode in the gigantic stipules 
peculiar to that genus. t 

* Les Nectaires, Ann. des Sci. Nat., Bot., vol. iii., p. 1, 1879; also, 
Etudes Anatomujues et Physiologiques des Nectaires, Compt. rend,, toni. 
Ixxxviii., p. 662, 1879; also, Cross and Self Fertilisation of Plants, ip. 402; 
also, Stadler, Beitr. %. Kenntniss d. Nectarieen u. Biologic d. Bliithen. 

t See Belt's Natui'alist in Nicaragua ; also a paper by F. Darwin, 
in Trans. Lin. Soc, on the same subject. 

Fig. 13. — Stipules of Impatiens : a section showing anatomy ; 
b, with a drop of honey in the centre (after Kerner). 



A microscopic examination of the anatomy of nectaries 
shows them to be composed of small cells closely resembling 
the merismatic condition of ordinary cellular tissue (see 
Fig. 43, a), and similar to the arrested parenchyma of the 
pulvinus at the base of the petiole of sleeping leaves, which 
enables that organ to remain flexible. Or, again, it is very 
similar to the conducting tissue of the style, which owes its 
origin to the irritating effect of the pollen-tubes (chap, xviii.). 

The function of the nectary is to secrete honey, or, to 
speak more accurately, either principally glucose, or else cane 
sugar, or both, for the proportion varies greatly.* 

The position of nectaries in flowers is very various, and any 
organs can form them. It will be enough to enumerate a few 
localities as follows: The Lime, species of Malpigliia,\ and 
perhaps Coronilla, furnish instances, which are comparatively 
rare, of the sepals of the calyx being 
nectariferous. In Buttercups, Hellebore, 
and Aconite, nectar is secreted by the 
petals or their representatives. In 
Violets, Atragene (Fig. 44), Fentstemon, 
and Stellaria the filaments undertake 
the duty, while in Caltha, Monotropa, 
and Rhododendron it is the carpels or 
pistil. In most instances the honey is 
secreted by glands, disks, etc., issuing 
out of the floral receptacle. If the ovary Fig. 44.— Petals passing into 
be inferior, then the secreting structure Saye\<?Str iimkr). ^^ 
is on its summit, as in the Umbelliferce ; 

and in that case it is the base of the styles from which the 
nectariferous tissue is developed. 

The Origin of jSTectaries. — Limiting one's self to those in 

* Bull. Soc. Bot. Fr., viii. (1886), Rev. Bill., p. 212. 
t Nature, vol. xvii., p. 78. 


flowers, there are many reasons for inferring their existence 
to be due to the direct and irritating action of insects them- 
selves when searching for juices as food or otherwise. 

That a merely mechanical irritation may cause a flow of 
nutrient fluid to the spot, so that the tissues may increase in 
size by the development of cells, which would not otherwise 
occur, is abundantly evident. It is seen, for example, in the 
growth and development of galls ; of the so-called " Ant- 
plants " on Myrmecodla (p. 115), Acacia sphcerocephala, etc.; 
in the thickening of all climbing organs as soon as the irrita- 
tion of the foreign body has commenced ; hence the inference 
that hypertrophy may occur wherever an insect's proboscis can 
irritate the floral organs, is by no means without foundation. 
Why the cell-contents of nectaries should especially give rise 
to sugar, is a question at present beyond answering. Those 
of conducting tissues appear to do the same. In the case of 
nectaries it may, perhaps, have originated as a pathological 
phenomenon which has become fixed and hereditary; for 
pathological conditions often determine a flow of gum, as in 
Cherry-trees, resins in the ConifercB, watery and sugary dis- 
charges from wounds, etc. ; and it is impossible to draw any 
hard-and-fast line between a pathological and varietal state : 
as, for example, in closing the scar after the fall of the leaf 
the fibro-vascular bundles are sometimes stopped by gum — 
a process which, in this case, might be regarded as normal, 
and not pathological as in the former. 

If a particular locality be perpetually irritated, so to say, 
for generations, all analogy shows that the effect may become 
permanent and hereditary ; at least, as long as the irritation 
is persistently renewed year after year. And, on the con- 
trary, the theory is equally supported by the negative evi- 
dence of the disappearance of the honey-glands whenever the 
whole flower degenerates and becomes regularly self- fertilising 


or else anemophilous. In these cases, in unison with, the 
degradation in size and colour of the corolla, or else its entire 
loss, the nectaries tend to and generally vanish entirely ; as 
may be seen in Polygonum aviculare as compared with P. 
Fagopyrum and P. Bistorta. 

The simple origin of nectaries, then, according to my 
theory, is that insects, having been attracted to the juicy 
tissues of flowers, by perpetually withdrawing fluids have 
thereby kept up a flow of the secretion which has become 
hereditary, while the irritated spot has developed into a 
glandular secreting organ.* These spots occur wherever the 
prevailing insect found it most convenient to search ; hence it 
is sometimes at one place, sometimes at another, even in 
closely allied plants. Thus, in Buttercups the stamens and 
carpels form a compact globe, especially the latter, and defy 
the penetration of a proboscis. The corolla, however, admits of 
an entrance of its base. In Atragene alpina the basal portion 
of the filament forms a nectary (Fig. 44). Comparing these 
with Caltha, the large carpels of this plant admit the passage 
of a proboscis between them; and the nectaries are now 
developed on the sides of the ovaries, exactly where they 
would be irritated. 

In Banunculus corttiscefolius, of 'the Canary Islands, which 
has a corolla more than two inches in diameter, the petals 
are entirely without honey-glands. On the other hand, the 
carpels are very large and flat, with plenty of space between 
them. Although I could detect no honey in plants grown at 
Floore, Weedon, the tissue over the centre of the ovary was 
modified, and exactly resembled the ordinary tissue of a 
honey-gland. If I am justified in assuming the carpels as 

* It is closely analogous to the action of the pollen-tube, which 
causes a flow of nutriment to the conducting tissue, only there is a 
physiological as well as mechanical irritation in that case. 


nectariferous, this would bear out the above remarks, for it 
would be as easily accessible as in the case of Caltlia. 

The merely occasional puncture and lesion caused by an 
insect which then flies away and does not keep up the irrita- 
tion — unless it be renewed by other insects — would not of 
itself be hereditary.* Thus, for example, Anemone nemorosa 
appears to be honeyless, but supplies pollen to bees ; yet 
Miiller noticed them frequently probing between the sepals 
and stamens, apparently to obtain juices wherewith to 
moisten the pollen. This process may have been the actual 
origin of nectaries, the result of a wound constantly repeated 
and kept up, being a flow of a sw^eet secretion, which has thus 
attracted insects and induced them to repeat the process. 

Analogous Cases. — A some^vhat analogous illustration is 
that of galls, but in them the presence of the egg, and sub- 
sequently the grub, keeps up the irritation. These remark- 
able structures do not form spontaneously as nectaries now 
do, without a puncture ; still, even in this case, there may 
be, for all we know, a predisposition to form them ; perhaps 
seen in the readiness with which the Oak forms so many 
kinds, and they may be now, perhaps, much larger than they 
were when insects of any particular species first punctured 
the ancestral oak upon which so many kinds have now been 
evolved. t The apex of a shoot of Yew attacked by Cecidomyia 
taxi is transformed into a fleshy ring curiously resembling 
the honey-disk of many flowers. 

It is well known that in the human subject there may be 
a predisposition for tumorous or cancerous growths which 
is hereditary; and there would seem to be a very close 

* Injuries, especially to the nerves, may be hereditary in man '. see 
Nature, xxiv., p. 257. 

t M. E. Heckel thinks that the female "gall-flowers" of the Fig, 
with an abortive ovary, in which the Cynips blastophaga lays its egg, is 
now an hereditary form (Bull. Soc. Bot. de Fr., 1886, p. 41). 


resemblance between tumours and galls, thougb. originating 
from different sources, both being liypertrophied conditions 
of certain normal tissues. For example. Sir B, C. Brodie 
thus describes a fatty tumour : " There is no distinct 
boundary to it, and you cannot say where the natural adipose 
structure ends and the morbid growth begins." It is pre- 
cisely similar with galls, which are due to cell-division 
setting in at certain points of the epidermis and subjacent 

Although lesions and mutilations will not as a rule prove 
to have any hereditary effects, yet the tendency to respond to 
an irritation becomes permanent, and the form and structure 
of the resulting organ may actually appear long before the 
irritation is applied. This is conspicuously the case in the 
tendrils of Ampelopsis Veitchii, in which the adhesive "pads" 
are in preparation before any contact with a wall has taken 
place. This is not the case with A. hederacea. Similarly the 
aerial and climbing roots of Ivy are regularly produced 
only on the shaded side. They can, however, be readily 
made to form on the opposite side, if that be artificially 
shaded ; and where, indeed, they may be not infrequently 
found in nature, where they can be of no use. Such cases 
prove that the tendency to produce them is an hereditary 
affection which is present before the irritation is brought into 
play. Again, with regard to the tendrils of the Cuctorbitacece, 
though the coiling does not take place till the irritating effect 
induced by contact with a foreign body has brought it about, 
yet the tendency is seemingly so strongly hereditary that, 
in several cases, the tendrils are coiled while undeveloped in 
the bud, and have to straighten themselves before again 
coiling on contact, as may be seen in the common Bryony. 

In the case, however, of a mutilation, when it has been 
ODce made, the place heals over, and there is an end of all 


special vital action at the place. If, however, the same 
place be induced to secrete bj constantly repeated irritations, 
as the same flower is repeatedly visited over and over again 
before it fades, and the flowers of its ofl^spring have to un- 
dergo the same process, year after year, generation after 
generation, I think it is at least a reasonable surmise that 
there will at last ensue a permanent flow of fluid to the place, 
with a corresponding modification of structure, and so the 
nectary becomes established. If, however, from any cause 
the flowers become neglected, then the nectaries degenerate 
and ultimately disappear. 

Apart from some general theory of the kind proposed, it 
is impossible to assign a reason for glands appearing at all 
sorts of places in flowers. A theory to be worthy of accep- 
tance must meet all cases, if possible, and I maintain that 
the one I propose is compatible with every observation that 
has been made in flowers.* 

* I would suggest a similar origin for the insectivorous pitchers of 
Nepenthes. They originate, as Sir J. D. Hooker has shown, from water- 
glands. The effort to dispose of water brought up by the fibro-vascular 
cord keeps the tissue of water-glands at the extremity of a cord in a 
state of plethora, thereby somewhat arresting any change of form and 
retaining the cells in the very characteristic merismatic stage. And if 
it now meet with an external irritation from insects attracted by the 
escape of fluids a further response to their influence begins, and the 
wonderful structures we are familiar with in the pitchers of Nepenthes 
are the final result. 

I see no greater difficulty in conceiving of such an origin than in 
any other complex structure, such as the human eye. If the latter could 
originate from an epidermal cell sensitive to light only, and by succes- 
sive increments, traceable more or less distinctly through the various 
strata of animal life, finally reach the highest and most complex form 
of that of man, there is nothing inconceivable in the growth and 
differentiation of a pitcher in response to an external stimulus. 

What I cannot conceive of is, that any organ has ever originated 
■without a definite stimulating cause acting persistently in one and the 


Witli reference to the continnous flow of nectar, I would 
draw some analogy from animal secretions. Mr. Darwin, in 
speaking of the cow, observes * : " We may attribute the 
excellence of our cows and of certain goats, partly to the 
continued selection of the best milking animals, and partly 
to the inherited effect of the increased action, through mans art, 
of the secreting glands.^' This fact, recorded in the last 
sentence, which I have italicized, is only one example of the 
general principle of increase of growth by use, which I take 
to be strictly analogous to what takes place in the vegetable 
kingdom. And we may notice, in its special application to 
the formation of glands and other structures by mechanical 
irritation, that it is none other than a mechanical irritation 
which keeps up the secretion of milk for prolonged periods. 

The common or physical basis of vegetable life, namely 
protoplasm, is very nearly f indistinguishable in its properties 
from that of animals. Their behaviour is every day being 
proved to be not only similar but identical in the two 
kingdoms. The effects, under mechanical irritations and 
strains, of nutritive matters of the same kind, of poisonous 
substances, of electricity, etc, all show that the bond which 
unites the animal and vegetable kingdoms together is of one 
and the same nature, and that the links of the chain are 
forged out of this common basis of life. 

It is not to be wondered at, then, but rather to be antici- 

same direction. In the case of the eye, I take that cause to be light. 
In the case of an irregular corolla or the pitcher of Nepenthes, I assume 
it to be insects {Tr. Lin. Soc, xxii., p. 415 j An7i. Sci. Nat., 4 ser., xii., 
p. 222). 

Conversely, in the absence of light the eye vanishes ; in the absence 
of insects, corolla, honey, etc., go ; so that negative evidence tends to 
support the positive in all cases alike ; see Or. of Sp., 6th ed., p. 110. 

* Anim. and PI, tender Dom., ii., p. 300. 

t See Journ. Roy. Micr. Soc. 1887, 771. 


pated, that tissues will behave alike in both kingdoms ; that 
organs will grow with use and degenerate with disuse ; that 
they will develop processes to meet strains put upon them, 
as the limbs of animals have done and as stems * will do by 
forming special tissues; and, on the other hand, that they 
will atrophy if not called upon to display their powers, as 
parasitic organisms abundantly show in both kingdoms ; and 
as plants degenerate in water, which saves them the trouble 
of supporting themselves. 

All this is exactly what one finds to be the case in every 
department of the animal and vegetable kingdoms alike, 
whenever we search diligently into the anatomy and meaning 
of the histological details of all parts of organisms. 

Correlations of Floral Nectaries with Pollination. — 
There is yet another point observable in glands. As the 
position of a gland or nectary is just where it is most easily 
accessible to the particular insects which visit the flower — 
a fact abundantly illustrated throughout the floral world, — 
and since the sole use of it to the plant, as far as we can 
see, is that it should attract insects which transfer the pollen 
from one flower to another, one naturally looks to see if 
the positions of the anthers and stigmas are in any way 
correlated to that of the honey-gland. Such is, in fact, 

* I would throw out a suggestion that the anomalous stems of 
climbers, which often develop supernumerary collateral axes, but all 
coherent in one common stem, may be due to a response to the strains 
to which those stems are subjected, occurring in various directions, as 
they hang dependent on other trees. Other peculiar features, as of 
innumerable vessels, feeble wood tissues, etc., I take to be due to 
degeneracy, through these stems not being self-supporting, so that they 
have assumed very much the anatomical characters of subterranean 
roots. Again, just as the pericycle plays so important a part in the 
structure of many roots, it will be found that this same active layer is 
the parent of at least several of the above-mentioned supernumerary 
tissues in climbers, as in the tendrils of Cucwrbita, Bryonia, etc. 


invariably the case; so that one cannot but infer tliat a 
common cause has brought about their correlated positions. 
This close correlation is, of course, especially observable in 
the more highly differentiated flowers. In regular flowers, 
accessible on all sides, the glands are placed symmetrically 
round the flower — whether on the sepals, as in Lime ; on the 
petals, as in a Buttercup ; or on the receptacle, as in Geranium 
pratense, — or else there is formed a disk, as in so many "disci- 
floral " plants. As soon, however, as a flower begins to show 
some tendency to irregularity, or the flower is visited in one 
way only, the honey-secreting organ at once becomes more 
restricted in localization; as in the Wallflow^er, where it forms 
two cushions, out of the middle of which the shorter 'stamens 
arise, while the petals form two pseudo-tubes leading down 
to those two glands. Again, in the Lahiatce, so markedly 
zygomorphic, the honey-gland is often restricted to the 
anterior side, on which the proboscis is inserted. Similarly 
in Antirrliinutn Tnajus, "the honey is secreted by the smooth 
green fleshy base of the ovary, whose upper part is paler in 
colour and covered with fine hairs ; ... it remains adherent 
to the nectary and to the base of the anterior stamens. The 
short wide spur permits the insect's proboscis to reach the 
honey from below ; above and in front it is protected by a 
thick fringe of stiff knobbed hairs on the angles of the 
anterior stamens."* 

It is hardly worth while giving other cases to prove the 
universal rule, that the position of the honey and its gland is 
always where it is most accessible; and the position of the 
anthers is, at the same time, just where they will be most 
conveniently struck by the insect; while the style and stigma 
supply a third correlation, so that the latter organ invariably 
hits the insect where the pollen has been previously placed. 
* Miiller, Fertilisation, etc., p. 433. 


One more point may be noticed in connection with the 
above-mentioned correlations, namely, the motility of many 
stamens. This is always in reference to fertilisation, and, 
if it be an adaptation to intercrossing, then the anther takes 
np such a position that the insect strikes it when searching 
for honey, as in the Aconite and Tropceolum. If, on the 
contrary, the motion is to secure self-fertilisation, then it 
is regardless of the honey, and may actually interfere 
with the access to it by insects, as in the Bosacece : for in 
members of this order, with an indefinite number of stamens, 
the further they spread away from the pistil the more readily 
is the honey accessible ; but when they curve inwards, and 
crowd over the stigmas in the centre, they completely cover 
up and conceal the honey-disk. 

The position of the anthers in relation to the honey- 
secreting organs wdll, I think, often be found to be the clue 
to certain anomalies in flowers. Thus in Geranium pratense 
it has been noticed that the petaline stamens stand ulti- 
mately externally to the calycine. Now, the position of the 
five glands in front of the sepals requires that a tubular space 
should exist above them, down which an insect may thrust 
its proboscis, as in the Wallflower. Consequently the five 
stamens in front of the sepals must be so disposed as not to 
interfere with this passage. This can only be secured by 
their bending well inwards towards the styles below, and 
then outwards, above, so as to bring the anthers again on 
the same vertical plane as those of the petaline stamens. 

The more internal position of the calycine stamens, and 
the external position of the petaline ones, are immediately due 
to the gland, so to say, forcing the former inwards, while 
the buttress-like bases of the carpels thrust the latter out- 
wards. This gives rise to the so-called obdiplostemony of 
the Geraniacece. 



General Illustrations — Protoplasmic Irritability. — Having 
now stated on what gronnds I believe that the cohesions and 
adhesions between them, as well as the forms of floral struc- 
tures have arisen — namely, in response to the ii^ritations set 
up mainly by insect agencies, coupled with the effects of 
nutrition, atrophy, hereditary influences, etc., — it will be 
desirable to show briefly, not only how remarkably sensitive 
almost all parts, both vegetative and reproductive, are to the 
action of stimuli, but how they exhibit even visibly respon- 
sive effects, both in the protoplasm of the cells and in the 
tissues which are composed of them. 

The sensitiveness of living protoplasm is one of its most 
marked and well-known phenomena. It exhibits changes in 
its distribution within the cell as well as motions, which are 
the direct result of external stimuli. These may be very 
various, such as light, heat, electricity, or a merely mechani- 
cal irritation, as well as organic and inorganic solutions. 

Of the effects of stimuli upon the protoplasm, some may 
be beneficial, and partake of the nature of nutiition, as may 
be witnessed in the protoplasmic " aggregation " of insec- 
tivorous plants.* Yery similar appearances follow electrical 

* See Darwin's Insectivorous Plants, fig. 7, p. 40. 



or mecliamcal irritations. Thus Fig. 45* shows the effect of 
electrical action on the threads of protoplasm ; a represents 
a cell of a hair of Tradescantia Virgi- 
niaca; h the same, after the application 
of an electrical cnn^ent. The following 
are Dr. Weiss's observations upon this 
phenomenon : — 

" A constant electrical current is 
without influence upon the protoplasmic 
excitation ; whereas the alternate shocks f 

* From Weiss's Anatomie der Pflanzen, p. 95. 

f Pfeffer has noticed that the weight, per se, 
"l"™!,?":',''™:! ^ot of the body in contact [if yery slight?] is of 
dition; b, under electrical no consequence to tendrils. Thus cotton-wool 
action Cafter Weiss). weighing "00025 grain produced no effect if 

carefully placed on them ; but it did when a gentle impact was caused 
by slight currents of air. Tentacles of Brosera have a sensitiveness very 
similar to that of tendrils, inasmuch as small splinters of glass only 
produced irritation of the glands when they caused a rubbing as the 
result of concussion (see Journ. Boy. Micr. Soc, 1886, p. 285). 

Pfeffer concludes that the conduction of sensitiveness is not alto- 
gether due to a continuity of protoplasm, as it does not extend to the 
epidermis. Since, however, the outer cell-wall of the epidermis can 
grow when in contact with a foreign body, it would seem to clearly 
indicate that under such circumstances it still retained its protoplasm ; 
and that the modern view of the cell-wall being at first a protoplasmic 
layer, and not altogether a dead secretion from the protoplasm within 
it, is correct ; for otherwise it is difiicult to imagine how it could adapt 
itself to the surfaces of foreign objects at all. 

Heckel, in studying the movements of the stamens in Sparmannia, 
Cistus, and Helianthemum, discovered that the epidermis plays an 
important part : " L'epiderme, contrairement a ce que voulait Morren 
(Ann. des Sci. Nat., t. xix., p. 104), est done dans quelques cas I'organe 
principal et visible du mouvement. Je me suis mieux assure du role 
qu'il remplit, en enlevant cet epidermq quand les dimensions des filets 
mobiles le permettaient sans mutilation profonde (Cistus ladaniferus) : 
tout mouvement alors etait suspendu " (Bull, de la Soc. Bot. de Fr., torn, 
xxi., 1874, p. 212). See below, p. 163. 


of an inductional apparatus always produce more or less deeply 
extending changes in the form of the plasm, which resumes 
its normal character if the power exerted be not too strong. 
The protoplasm immediately forms itself under the induc- 
tional shocks into lumps or balls ; and, moreover, often sends 
club-shaped extensions with great suddenness and energy 
into the cell lumen, and immediately brings the circulation to 
a standstill. The rotation returns, however, after a period of 
rest, the extensions are drawn in, and the former net-shaped 
distribution of the protoplasm is restored, even when the 
whole mass of the plasm has been changed into a number of 
colourless balls and lumps. If the current is allowed to go 
only through a limited portion of the cell, then the streaming 
stops also in this tract only, and that, too, amid the formation 
of the lumps and balls. 

" A sudden increase and decrease of temperature acts in 
the same way ; there ensues a formation of drops, a cessation 
of the current, etc. Yet even here a return to the normal 
constitution takes place if no real coagulation of the proto- 
plasm has occurred. On the contrary, the current often ensues, 
after its recommencement,' with a greatly heightened speed, 
and even boisterously. The grains, etc., found in the cell-sap 
outside the protoplasm are, however violently the current 
may flow, in no way influenced by it, but remain at rest.'* 

M. E. Heckel has described * the effect of a mechanical 
irritation on the protoplasm of the cells of the filaments of 
Berberis. He says that the cells of the irritable part are 
arranged in a parallel manner, being longer than broad. 
Their contents are of a yellow colour, and disseminated 
throughout the whole cavity, but especially applied upon 
the walls. After receiving the excitation, the same cells, 
the surface of which is striated transversely, are massed 
* Bull, de la Soc. Bot. de Fr.y torn, xxi., 1874, p. 208. 


together or aggregated, so as to occupy only two-thirds of 
the space they formerly required. The contents, retreating 
from different points of the circumference, are condensed in 
the centre of the cell, and the transverse strise are pro- 
nounced in a high degree. The cells at the back of the fila- 
ment are contracted in repose, and extended under irritation, 
i.e. in an opposite manner to that of the other side of the 
filament. The irritability does not reside in the epidermis. 

A result of this aggregation must be a frequent displace- 
ment of the nucleus. In Weiss's figure the irritation hap- 
pened to be made apparently at one end of the cell, while the 
nucleus was at the other ; but in Heckel's description it 
appears that the protoplasm is drawn from every point ; 
so that, supposing the nucleus had been at the lower end 
of the cell (Fig. 45, a), it would have been most probably 
displaced. The consequence would be, that if such a nucleus 
formed its cell-plate, the ultimate position of that plate 
would be different from what it would have been had no 
irritation been applied to the organ. 

Though one does not look to electricity as a cause in 
nature, yet that light determines the direction of cell-division 
is abundantly proved in the case of leaves, whose tissues alter 
according to their position ; the palisade cells, for instance, 
bring formed on both sides, if the exposure to light be equal, 
or on the under side if that be placed uppermost. Similarly 
does it influence the formation of stomata.* Again, Stahl has 
shown that the direction of the division of the nucleus, 
which takes place in the spores of E^uisetum depends upon 
the direction of the rays of light ; the two daughter-nuclei 
lying in the direction of the o-ay. On the other hand, the 
nucleus at the greater distance from the source of light is 
that of the root-cell, while the one nearer to the source of 
* M. L. Dufour., Ann. Sci. Nat., torn. 50 (1887), p. 33 1. See below, p. 173. 


light is that of the prothallium cell.* Climbing roots of 
Ivy also appear on the darker side of the shoot, etc. 

It is impossible to regard the above cases as isolated, but 
they are special instances, revealing not only the general 
irritability of protoplasm, but the minuter effects upon the 
nucleus, which, in its turn, is thus compelled " to respond," 
and sets up cell-division, i.e. the formation of a tissue in the 
direction of the external influence, as mentioned above in the 
sentence I have italicized. 

The next very important point to notice is that cell-divi- 
sion can take place in response to, and in the direction of an 
external mechanical stimulus, just as well as in that of light. 
As the sensitive plant is influenced by, and visibly moves its 
foliage under the irritation of a touch or of varying degrees 
of light, so do I assume that the peculiar anatomical 
structures which permit of those motions are the direct result 
of external stimuli. Sjparmannia, it may be added, exhibits 
three kinds of movement, viz., Sleep in the calyx and corolla, 
mechanical irritahility in the stamens, and an elevation of the 
peduncle. (See Heckel, I.e., p. 210.) If this position be 
granted we have at least a working hypothesis for the 
present theory of the origin of floral structures. 

Formation of Tissues due to Irritability. — Apart from 
the preceding theoretical supposition, there may be fre- 
quently witnessed an actual formation of tissues of various 
kinds, through hypertrophy on the one hand, often coupled 
with atrophy on the other, and entirely brought about by 
physical or mechanical irritations. Cell-division is thus set 
up, a result which would not have occurred had not the 
external stimulus been applied. 

It is an important fact to notice, that in some cases the 
abnormal growth, though immediately following the stimulus, 
* See Jl. Roy. Micr. Soc, 1886, p. 287 ; and Bull. Soc. Bot. Fr., 21, p. 65. 


and never occurring without it, leaves no hereditary effect as 
in the case of galls * and of the thickening of the tissues of 
some climbers after they have caught and clung to a foreign 
body, such as the petioles of Clematis,^ and the hooked 
peduncles of Uncaria (Fig. 46). In other cases the effect 

has become hereditary, and may 
then be regarded as a specific 
character. These diff'erences are 
well seen in the tendrils of Ampe- 
lojpsis Jiederacea as compared 
■with those of A. Veitchii. In 
the former there are no traces of 
the adhesive " pads " at the ter- 
minations of the slender hooked 
tips of the branching tendrils, 
Fig 46 -Climbing peduncle of rncana, Until contact with the surface of 
(after'Trtubt" '"''^'"^ ^ '"''^''' ^ Wall has occurrcd. On the 

latter species, however, the pads 
are in course of development before any contact has taken 
place just as the aerial roots of Ivy begin to appear before 
contact. It is therefore reasonable to conclude that the 
effect of contact has become more or less hereditary in 
the latter Japanese species, though not in the American. 

These tendrils behave exactly like the clasping roots of 
Orchids, Ivy, etc., as well as the so-called " roots " of Lami- 
naria, Cutleria, etc. Indeed, the way in which subterra- 
nean root-hairs fix themselves to particles of the soil is by 
essentially the same method. The irritation caused by con- 
tact aided by moisture excites the cell- wall to grow out into 
protuberant processes, which enables it to adapt itself to the 

* I have examined a considerable number of galls, and can quite 
corroborate M. Prillieux, who has shown how the normal tissues become 
hypertrophied {Ann. des Sci. Nat., ser. 6, tom. ii. (1876), p. 113). 

t See Climbing Plants, fig. 1, p. 47, and fig. 4, p. 74. 


irregularities of the surface of the particles. An excretion 
of mucilage appears to follow, which fixes the organ to the 
foreign support. The irritation not only affects the epider- 
mal layer, but the subjacent tissues as well, which then assist 
the former in grasping the support.* 

Another result of growth due to external agencies is seen 
in the hypertrophied stipules of Acacia sphcerocephala and 
the stems of Myrmecodium, etc., in consequence of the irritation 
set up by ants. Dr. Beccari f (and M. Treub J) has examined 
these " Ant-plants," which occur in Uuhiacece, Mijristicacece, 
Eupliorhiacece, Verhenacece, Melastomacece, and Palmce, and 
explains the abnormal structures by variability and heredity. 
A small swelling appears on the tigellum of Myrmecodium 
serving the purpose of a reservoir of water, but which only 
grows larger through the agency of ants. These creatures 
induce hypertrophy of the cellular tissue. This, then, be- 
comes hereditary. I would venture to go further, and 
attribute the large honey-pits at the base of the leaf-stalk on 
Acacia sphcerocephala, as well as the terminal " fruit-bodies " 
occurring on the tips of the leaflets, to the same cause, viz. 
the mechanical irritation of the ants. 

There is, in fact, an abundance of evidence to prove that 
many organs of a plant, if subjected to irritation, can respond 
to it, and not only increase in size by hypertrophy, but 
materially alter their anatomical structure and develop new 
processes. Secondly, that these altered states, if the irrita- 
tion be persisted in, may become hereditary. 

* See Fig. 42, a, (p. 137), which represents the inferior side of an 
aerial root of PhalcBnopsis amahilis in contact with a surface ; h is that 
of a root which has penetrated the soil (Organisation dorsiventrale dans 
les Bacines des Orchidees, par M. E. Janczewski. Ann. des Sci. Nat, 
ser. 7, torn, ii., p. 55. 

t Malesia, ii. (1884). See Arch. Ital. de Biol., vi. (1885), p. 305. 

X Ann. Jard. Bot. Buit, iii., p. 129 (1882). 


The influence of the environment upon the anatomical 
and morphological structures of plants has been lately and 
widely studied from several points of view ; and it has been 
shown conclusively, by Constatin and others, how a change of 
medium — as, for example, from air to a subterranean one, or, 
again, to water — profoundly affects every tissue of the plant, 
whether the root, stem, or leaves be submitted to it. So, too, 
leaves of many plants have been proved to be very sensitive 
to changes of position and to different amounts of light — 
which is a most potent and exciting cause in affecting the 
mesophyl, palisade, and other tissues, including the epider- 
mis, stomata, and even cuticle. It is foreign to my purpose 
to describe or discuss these details in the vegetative system 
of plants ; my sole object being to draw attention to the fact, 
and then to apply it to the structure of flowers. 

Ieritabilitt of the Floral Organs. — Perhaps no parts of 
plants are more keenly sensitive to stimuli, or show a greater 
number and variety of results to excitement than flowers. 
A large proportion resemble plants which sleep, i.e. they 
exhibit movements according to the amount of light and heat 
which they receive. So various is this, that Linnasus was 
able to frame his floral clock. While many thus open their 
petals at definite periods and subsequently close them and 
die, as Convolvulus ; yet a large number reopen them again 
when the due amount of light returns, like Daisies and 
other Composites, AnagalUs arvensis, Mesemhryanthemum, etc. 
Others, like Silene nutans, unroll their petals at night, but 
roll them up again by day.* Besides these spontaneous 
motions of the perianth, the stamens often exhibit move- 
ments, apparently wnthout any external stimulus. Thus 
Parnassia and Saxifrages slowly move their stamens in suc- 

* See Dr, Kemei^s descrip]tio^ of tliis flower. Flowers and their 
Unhidden Guests, p. 133. 


cession, either towards tlie pistil as in the latter, or away 
from it as in the former. Other flowers, like Cratcegus, RuhicSj 
and AUsma, have them at first spreading away from, but 
afterwards bending over the pistil. These processes facilitate 
one or other kind of fertilisation, and are very common. 

Slow movements of the filaments after the anthers have 
discharged their pollen, so as to place them out of the way 
of the pistil, are not at all uncommon in strongly protan- 
drous flowers. Echium* and Teucrium Scorodonia'\ will illus- 
trate this well-known phenomenon. The "lemon-scented" 
or " oak-leaved " species of Felargonium has small and very 
irregular flowers, somewhat papilionaceous in appearance, 
with the stamens declinate, lying on the anterior petal ; the 
style lies beneath them, with the five stigmas quite undeve- 
loped. After the anthers have shed their pollen, they fall 
off, and the filaments bend down outside the flower, while 
the stigmas now come to maturity and lie in the very place 
where the anthers lay before them. 

Similar slow movements are very common in the styles 
and stigmas of plants. In the Compositce and Campanula, 
Lohelia, Gentiana, etc., the style arms with their stigmatic 
papillae curl backwards, and so secure self-fertilisation. 
In several of the Scropliularinece and Labiatce, the style 
gradually bends over, so that the stigma comes in contact 
with the pollen. This, however, may be partly due to pro- 
longed growth. As examples, may be mentioned Hhinanthus, 
Melampyrum, Galeopsis, Stachys sylvatica, etc. Treviranus 
says the same thing occurs with Gladiolus, the style curving 
back towards the anthers. J 

* Cf. rigs. 20, 6 and c, p. 82 : & shows the position of the stamens 
before polKnatiou ; c, after it. 

t See Miiller's Fertilisation, etc., p. 500, fig. 169. 
X Ibid., p. 548. 



In addition to slow and seemingly spontaneous move- 
ments, to which all organs of a flower are liable, there are 
many rapid actions, brought about by the direct means of 
external stimuli applied to them. Thus Indigofera and 
Genista are two genera in which the claws of the petals are in 
a great state of tension when the flower is open, and the 
moment they are touched it explodes. The claws, from 
having been horizontal, curl down- 
wards, and tbe staminal tube with 
the included pistil is jerked up- 
wards. Thus Fig. 47, a, represents 
a flower of G. tinctoria just expanded. 
On passing a pencil point down the 
front of the standard, the wings and 
keel petals drop vertically, as seen 
in Fig. 47, 6, looked at from the 
front. The staminal tube now lies 
against the standard. The keel, 
from its extreme tension, splits 
where it curls at the base, and be- 
comes wrinkled in front, as seen in 
Fig. 47, c. 

There is a plant of the order 
Convolvulacece, the corolla of which 
„. ,^ ^ ,, ,. , . . actually closes on receiving a me- 

fore. ?>. after explosion; c, claws chanical toUch. M. H. Dutrochct, 

of keel. 

after observing that the movements 
of Mimosa pudica and Dioncea muscipula are all in one 
direction only, as also of the stamens of Cactus opuntia and 
Berberis, adds : *' Mais il est quelques cas oil cette incurvation 
oscillatoire s'effectue dans plusieurs sens diffcrents, tel est, 
par exemple, le phenomene que presente une plante du genre 
Ypomcea, observee aux Antilles par M. Turpin, plante encore 


inedite, qu'il designe sous le nom d'Ypomcea sensitiva. Le 
tissTi membraneux de la corolle campanulee, de cette plante 
est soutenu par des filets on par des nervures qui, au moiudre 
attouchement, se plissent ou ^'incurvent sinueusement, de 
maniere a entrainer le tissu membrane iix de la corolle, 
laquelle, de cette maniere, se ferme completement ; elle 
ne tarde point a s'ouvrir de nouveau lorsque la cause qui 
avait determine sa plicature a cesse d'agir." * M. Dutrocbet 
tben observes tliat tbis pbenomenon is in no way essentially 
different from tbe closing of tbe corolla of Convolvulus, to 
wbich Ypomcea is nearly allied, wben it passes into tbe sleep- 
ing state, as does tbe calyx or periantb of tbe Nyctagineoe. 
Lopezia coronata exbibits a curious and rapid movement 

a i c 

Fig. 48. — Lopezia (after Hildebrand). (For description, see text.) 

in a staminode. Miiller tbus describes it : f "In eacb flower 
tbere is present one perfect stamen ; a second, standing 
immediately below, is reduced to a spatbulate leaf, wbose 
two balves fold upwards, and, in tbe first stage, projecting 
borizontally from the flower, inclose tbe antber of tbe perfect 
stamen (Fig. 48, a). Tbe stalk of tbe spatbulate leaf bas an 
elastic tension downwards (b) ; tbe filament of tbe stamen 
an elastic tension upwards (6), so wben an insect aligbts on 
tbe projecting spoon-sbaped blade, as tbe only convenient 

* Recherches Anatomiques et JPhysiolngiques sur la Structure Intime 
des Animaux et des Vegetaux et sur leur Motility, 1824, p. 64. 

t Fertilisation, etc., p. 265. 



spot from which to reach two drops of honey that seem to 
rest upon a knee-shaped bend in the upper petals (a), the leaf 
springs downwards (fe), and the stamen is set free and flies 
upwards, dusting the lower surface of the insect with pollen. 
When the stamen has thus served its purpose, it gradually 
curves upwards out of the flower (c), and the style which was 
hitherto undeveloped grows gradually out of the flower in a 
horizontal direction, so as to form another alighting place (c)." 
Rapid movements in the stamens are not unknown. 
I described that of Medicago * many years ago, and now 
supply figures. Fig. 49, a repre- 
sents the front view of a flower on 
expansion ; 6, the same after a bee 
has exploded it — the staminal 
column has now arisen, curled up- 
wards, and abuts against the 
standard; c shows the curved posi- 
tion of the stamens, the corolla 
being removed. The stamens are 
inelastic, as they will not return to 
a horizonal position without break- 
ing across, if pressed downwards. 
Many other rapid movements of the filaments are too well 
known to need description, such as those of Berberis, Helian- 
themiim, Sparmannia, Centaurea, and TJrtica ; while Orchids 
exhibit various movements in the caudicles of their pollinia. 

Besides slow movements, the pistil often exhibits rapid 
ones on being touched, as are known to occur in Stylidium, 
Canna, Mar ant a and allied plants ; while the flap-like stigmas 
of Mimulus^f and of several genera of orders allied to the 
Scrophularineo}, close together on being irritated mechanically. 

* Journ. Lin. Soc, vol. ix. p. 327. 

t Mr. F. W. Oliver has lately investigated the mode of conduction 

Fig. 19.— Medicago sativa. (For de- 
scription, see text ) 


There is no need to describe a long series of movements, 
my object being simply to emphasize the fact that sensitive- 
ness and irritability are pronounced phenomena in flowers, 
which point to a highly irritable condition of the protoplasm 
contained in the cells of all the floral members.* And, although 
we cannot now trace the progress of change in the floral 
organs under the mechanical and physiological impulses due 
to insect agency, the probability that these have been the 
actual influences to which the tissues have responded, and 
thence evolved the existing floral structures, will now, I trust, 
appear to the reader to be of a very high order. 

of the irritation in the stigmas of Martynia lutea and M. prohoscidea, and 
of Mimulus lufeus and M. cardinalis. He believes it to be due to the 
continuity of the protoplasm from cell to cell. The tissue of the stigma 
consists of two lamellae. The irritability is confined to sevei*al layers 
of prismatic cells on the inner side of the lamella, where the continuity 
of protoplasm was determined. (Quoted from Journ. Boy. Micr. Soc, 
1887, p. 781. Ber. Deutsch. Bot. Gesell., v. (1887), p. 162.) 

Mr. Oliver has also lately contributed a valuable paper to the Annals 
of Botany (vol. i., p. 237, pi. xii., 1888), on " The Sensitive Labellum of 
Masdevallia muscosa." Continuity of the protoplasm occurs in the irritable 
"crest" on the labellum, which rapidly rises on being touched; the 
mechanism being closely comparable with that of the pulvinus of Mimosa. 
The author corroborates Mr. Gardiner's observation that a large amount 
of tannin occurs in the cells with which such irritability is concerned. 
References are also given to descriptions of other Orchids remarkable 
for having irritable perianths. 

* For further information on the effects of light and heat upon the 
opening and closing of flowers, the reader is referred to Sachs' 
Physiology of Plants, chap, xxxvi., p. 641, where the author gives an 
account of Pfeffer's investigations. It is not clear, however, how 
temperature acts. A casual discovery may perhaps supply a hint. On 
forcing air into the flower-stalk of the white Water-lily, I found that the 
petals instantly spread open. May not, therefore, a rise of temperature 
cause the air within the tissues to expand, and so at least help to produce 
the same effect ? 





The first effect produced by the action of the germination 
•of the pollen- tube is the formation of the so-called conducting 
tissue or layers of specialized cells which nourish the tube 
in its downward growth. Like glandular nectaries, this 
tissue consists of small merismatic-like cells, highly charged 
with nutritive and saccharine substances. In some cases it 
is a metamorphosed condition of the epidermis alone, as 

Fig. 50.- 

section of (epidermal) conducting tissue of Fumaria; h, that of Rubus; 
c, section of ovary of Crucifer (after Capes.) 

M. Capes has shown in his researches,* as in Fumaria. Fig. 
60, a, represents a section of the stylar canal, the lining 
epidermis having its cells charged with such matters, while 

* Ann. des Sci. Nat, vii., 1878, p. 209. 


three pollen-tubes are seen in section. Fig. 50, 6, shows the 
formation of conducting tissue at the angle of the inflected 
carpellarj edges of Buhus. The epidermal and subjacent 
cells form the conducting tissue in this case. The cells on 
the outskirts are charged with sphaera.phids. Fig. 50, c, is a 
section of the ova-ry of a Crucifer. The replum or false dissipi- 
ment, as in the Papavei'acece, forms the machinery for conduct- 
ing the tubes. The dotted lines show the original lines of 
fusion. Now, if my theory be true, that no structure exists 
which has not been brought into existence through some 
foreign action having been brought to bear upon it — either 
directly from without, as insect agency, light, etc., or 
indirectly through nutrition within the plant, — then, the 
existence of this specialized tissue would never have arisen 
had it not been for the iiTitating action of the pollen-tubes. 
The analogous influence of the mycelium of a parasitic 
fungus here gives us the clue. As such causes hypertrophy 
to set in, and induces nutritive matters to accumulate upon 
which the fungus lives, — just as the irritation of the egg or 
pupa of a cynips or other insect causes a similar accumula- 
tion of richly nutritive substances to be made within the 
tissues of the gall upon which it feeds, — so the germinating 
power of the pollen-grain and the growth of the pollen- 
tube have actually brought about the formation of these 
highly nutritive conducting tissues of the style. The effect 
has then become hereditary, so that they are now in course 
of formation, at least, during the development of the flower 
in preparation for the ingress of the pollen-tubes. 

The remarkably stimulating action of the pollen-tube had 
been observed more especially in Orchids. Hildebrand 
noticed that the influence of the pollen was twofold, in that 
it determined the growth of the ovary and the complete 
formation of the ovules before the process of fecundation had 


taken place.* M. Guignard has described the effects result- 
ing from his experiments.! Thus, in the case of Vanilla 
aromatica^ he found the development of the ovary was very 
rapid after poUinisation. At the time of flowering, the 
placentas have only the rudiments of the papillae which 
will develop into ovules, and the conducting tissue formed 
by the epidermis and subjacent layers on either side of 
the placentary projections is still undifferentiated. In 
the intervals which separate the bands of conducting tissue, 
corresponding to the midribs of the carpels, there is no 
appreciable modifications before fecundation ; but as soon 
as that has taken place, a layer of elongated papillae, filled 
with a granular substance, arises. With regard to the 
development of ovules, M. Gruignard remarks : " La poUinisa- 
tion et la germination du pollen sont indispensables a leur 
formation. L'ovaire d'une fleur non pollinisee ne s'accroit 
pas et tombe quelques jours apres I'epanouissement." 

As soon, however, as the pollen-tubes are formed, the 
ovules begin to grow, until the twentieth day, when the pri- 
mine thickens (much more than in other orchids) and finally 
gives to the matured ovule a globular form. 

In the mean time the embryo-sac and sexual apparatus 
have been forming, and are completed (excepting the fusion 
of the two members of each tetrad, which does not take place 
to form the secondary embryo-sac nucleus) in little more 
than a month after poUinisation. Five weeks after that 
period, fecundation commences. 

In following the progress of the pollen-tubes, it is not 

* Die Fruchthildung der Orchideen, ein Betveis fur doppelte Virkung 
des Pollen, Bot. Zeit., 1863. Bastardirungsversuche an Orchideen, Bot. 
Zeit., 1865. 

t Ann. des Sci. Nat., 1886, torn, iv., p. 202; see also Maury, 
Observations sur la Pollinisations des Orchidees, comp. rend, de I'Acad. 
des Sci., 2 Aout, 1886 j and also Guignard, do., 19 Juillet, 1886. 


till from tlie twelfth to the fifteenth day that some of them 
arrive at the base of the ovarj. Before the sexual apparatus 
is complete, the extremities of the pollen-tubes separate from 
the mass of tubes overlying the conducting tissue, twist in 
various directions, scrambling over the placentary lobes and 
their ramifications, and so approach nearer and nearer to the 
ovular fucicles ; but they only penetrate the micropyles, 
after the formation of the sexual apparatus. It is supposed 
by Strasburger that the synergid^ expel a liquid destined 
to guide the pollen-tube to the embryo-sac ; others think their 
function is to aid in the solution of tissues for nourishment. 
In Vanilla, for example, the upper part of the embryo-sac 
is absorbed where occupied by the synergidae, and is then 
covered by the elongated border of the primine. M. Guignard, 
however, adds : — " II est possible qu'il soit attire par un 
liquide expulse par les synergides,* comme le pense M. Stras- 
burger, ou bien aussi, comme je crois I'avoir constate, par 
I'etat special de la couche superficielle des membranes cellu- 
laires da bord interne du tegument." f 

On the action of the pollen-tubes M. Guignard writes as 
follows: — "Au contact des faisceaux polliniques, le tissu 
conducteur offre un contenu riche en sucre reducteur ; I'ami- 
don, dans le cas actnel, ne se trouve qu'au voisinage et du 
cote externe des faisceaux libero-ligneux des parois ovari- 
ennes. Outre le pouvoir d'attaquer la substance amylacee et 
d'intervertir la saccharose, comme I'ont montre tout recem- 
ment M. Yan Tieghem, J et M. Strasburger, § les tubes polli- 
niques peuvent aussi, a I'aide des ferments qu'ils contiennent, 

* Synergidce is better, being nearer Sunergatai. 
t L.c, p. 209. 

X Sur V Inversion du Sucre de Canne jpar le Pollen, Bull. Soc. Bot. de 
France, 1886. 

§ Ueher Fremdartige Bestauhung, Pringsh. Jahrb., vol. xvii. 


dissoudre la cellulose, aiusi que le prouvent les soudures avec 
fusion que j'ai observees plusieurs fois entre eux dans les 
cultures, oil le phenomene est plus facile a voir. D'ailleurs, 
la penetration directe des tubes polliniques dans les papilles 
du stigmate de plusieurs fleurs, apres dissolution de la 
membrane cellulaire, est un fait du meme ordre." 

I quote this passage in full, tbat the reader may see bow 
it completely corroborates my belief that the metamorphosis 
of the epidermis and subjacent layers to form the conducting 
tissue is entirely owing to the action of the tubes themselves, 
as well as the conversion of starch into saccharine, and there- 
fore easily absorbable matters. 

M. P. Maury has noticed very analogous facts in Verhas- 
cum, in that " at the period of pollination the ovules are still 
in a rudimentary condition, and altogether unfit for fertilisa- 
tion. The nucellus is entirely occupied by the embryo-sac, 
in the protoplasmic contents of which there is as yet no 
differentiation of oosphere, synergidae, or antipodals. It is 
only after the pollen- tube reaches the micropylar canal that 
these begin to be formed." * 

This observation corroborates what I have said above, 
that not only is the pistil delayed in development in insect- 
crossing flowers, but that arrest of growth may affect all 
parts, and particularly the ovules ; and I strongly suspect if 
more instances, of the Gamopetalce especially, were examined 
it would be found to be the rule and not an exception. M. 
Maury's investigations also agree with M. Gruignard's, in 
that the action of the pollen-tube is a stimulating one, and 
brings about developments which would not, and, indeed, 
cannot, otherwise take place. 

In Vanda tricolor pallens, experimented upon by M. 

* Bull. Soc. Bot. Fr., viii. (1886), p. 529, quoted from notice in 
Jov/m. Roy. Micr. Soc, 1887, p. 433. 


Guignard, lie noticed the not infrequent effect of a rapid 
change of colour in the perianth after pollination, although 
it did not fade for a week. The swelling began on the 
second day ia the " gynosteme," and progressed towards the 
ovary. From having been four centimetres long on the day 
of pollination, December 4th, 1885, by the 15th of April, 
1886, it had grown to seven centimetres. The ovules, how- 
ever, w^ere not full grow^n, the embryo-sac having still its 
primitive nucleus ; by the 15th of May, the ovules had 
attained their complete development. By the 1st of June, 
fecundation had taken place in nearly all the ovules. Hence 
about six months were required for the process. 

In this species the spaces over the mid-ribs were covered 
with long hairs, corresponding to the papillae in Vanilla. In 
both they appear to have grown after, and as a result of, 

In a flower ol Angnecum superhum w^hich became arrested 
the influence of the pollen-tube was remarkably illustrated. 
Three weeks after pollination an arrest of development 
followed in the ovary ; it had sensibly increased in diameter 
in the upper part. On examining the ovarian cavity at the 
top, M. Guignard found only a small number of pollen tubes, 
relatively short in length. f 

Another abnormal case was a Vanilla, in which, from 
some unknown cause, only two bundles of pollen-tubes were 
formed on either side of a placenta. Here the ovary grew 
on that side, causing a strong curvature. On the opposite side, 
the wall and the placentas with their ovules were atrophied. 

* Max Wichura found that silky hairs were sometimes the sole 
result of his attempts to hybridize willows ; and as analogous instances 
are the clothing the interior and exterior surfaces of galls with papillae 
or hairs, an indirect result of the irritation set up by the pupae (p. 138). 

t A like interpretation may be given to Vegetable Marrows when 
they swell only at their distal end. 



The exciting effect of the tubes is seen when Orchids are 
crossed which have no affinity, and are therefore incapable 
of fertilisation. Thus, the pollination of Orchis mascula by 
Cijpripedium paviflorum even determined the formation of 
the sexual apparatus in the former. Similarly, when Orchis 
and Listera, as well as Ophrys and Limodorum were crossed, 
ovules reaching various degrees of development were pro- 
duced, but none were impregnated. 

Everything indicates (writes M. Guignard) that the 
development of the ovules is subordinated to that of the 
ovary. In exotic Orchids the thickness of this organ and its 
elongation are often very pronounced before the appearance 
of the ovules.* 

Analogous results have been obtained by Max Wichura's 
experiments f in hybridizing Willows, who noticed the 
following degrees of failure indicating the various amounts 
of influence that the pollen-tube had over the sexual ap- 
paratus of the plants crossed : (1) the ovaries swell and 
ripen, but contain no seed ; (2) the ovaries are quite filled 
with silky hairs which clothe the umbilical cord end of the 
seed, but contain no embryo ; (3) seeds are present, but 
small and incapable of germination ; (4) seeds apparently 
perfect, but do not germinate ; (5) seeds germinate, but the 
seedlings are weak, and soon wither ; (6) seeds few but fertile 
and active ; (7) seeds numerous with only a few fertile ; (8) 
seeds numerous and fertile. 

* Gaertner, in his Memoire sur les Organes Reprodudeurs des Phaniro- 
games, devotes a special chapter to the enlargement of the ovary without 
previous pollination, with the result of a pseudo-fruit {Versuche u. 
Beoh. uber die Befrucht. Organe der Vollk. Gewdchse, 1844). 

t Die Bastardhefruchtung in Pflanzenreich, erldutert an den Bastarden 
der Deiden, Von Max Wichura. Mit zwei Tafeln. 4to., Breslau, 1865. 
Abstract by Rev. M. J. Berkeley, Journ. Roy. Hort. Soc, New Series, 
vol. i., p. 57. 


Similar results occur in many cultivated plants without 
hybridizing; as appear in seedless Oranges; Grapes, such as 
*' Sultanas ; " Bananas, Cucumbers, etc. 

Every other cause capable of acting in the same way will 
produce a like result, as in various instances of parasitism, 
when the cells become hypertrophied, as do those occupied 
by Synchytrium, or as in the roots invaded by Plasmodiophora. 
M. Guignard quotes an interesting case, which fully bears 
out the theory advanced in this book, of the results of the 
irritation of insects. He says, " A I'appui de cette maniere 
du voir je citerai une observation interessante que le hasard 
a fournie a M. Treub,* et dont ce savant a bien compris la 
signification reelle." 

" Ay ant rencontre des ovaires de Liparis latifolia qui 
presentaient un epaississement plus ou moins considerable, 
meme dans les fleurs non epanouies, et ou la poUinisation 
directe ou indirecte n'avait pu se faire, il trouva a I'interieur 
des petites larves qui y avaient penetre de tres bonne heure. 
Ces larves ne paraissaient exercer aucune influence nuisible 
sur les cellules, et semblaient avoir la faculte de se mouvoir 
librement dans la cavite ovarienne, bien qu'on les trouvat au 
contact de la paroi ou des placentas. Elles se nourrissaient 
evidemment des sues de I'organe envahi ; a peine voyait-on 
nne legere alteration de quelques cellules avec lesquelles elles 
etaient en contact. Compares a ceux des fleurs normales 
avant la poUinisation, ces ovaires habites par les larves 
olf raient des placentas plus grands et plus digites, sur lesquels 
s'etaient developpes finalement des ovules revetus de leurs 
deux teguments formes comme sous I'influence de la poUi- 
nisation, Les dimensions des ovules ne differaient pas de 
ceux des graines mures provenant d'ovaires pollinees, et non 
envahis par des larves. 
* Notes sur VEmhryon, etc., Ann. du Jard. Buit., iii., p. 121, pi. xix. 


" II etait done evident que les parasites avaient determine 
les memes effets que les tubes pollmiques : raccroisement des 
ovaires et des placentas et le developpement des ovules." 

The reader will here see the importance of this curious 
instance as bearing upon my general theory of growth in 
response to irritation ; so that if ovaries, placentas, and ovules 
can be stimulated into growth and developraent, there is 
no a priori reason why other parts of flowers may not equally 
well grow in response to irritations set up by the insect 
visitors ; as I have already shown to be the case in Clerodendron* 
and in Mr. O'Brien's experiments. t 

Perhaps it will not be amiss to notice here a very similar 
action of the suspensor in Orchids, described by M. Treub, 
which grows "backwards," escapes from the micropyle, and 
then ramifies in various ways, clasping and burrowing into 
the ovarian walls like a parasite in order to convey nutritive 
matters to the rudimentary pro-embryo. + 

Finally, M. Guignard remarks upon the degradations in 
the essential organs of Orchids as accounting for the well- 
known difficulty in raising seed from them : " Malgre le 
nombre immense des grains formees dans les conditions 
naturelles comme dans les serres, nombre qui parait etfe 
d'ailleurs une signe de degradation physiologique dans une 
famille oii la differenciation morphologiqne des organes floraux 
est cependant si elevee, I'insuffisance de reserve alimentaire 
contenue dans leur embryon microscopique, en necessitant des 
conditions speciales pour le developpement, suffit peut-etre a 
expliquer les difficultes et les insucces de la reproduction des 
orcbidees par gi'aines, et la parcimonie relative avec laquelle 
elles sont distribuees dans la nature." 

* See p. 130. t See p. 114. 

X Notes sur VEmhryog^nie de quelques Orchid^es, Verhandelingen der 
Koninklijke Akadamie van Wetenschappen, 1879. 


Witli regard to the difficulty of rearing Orchids, the 
reader may be referred to the Report on the Orchids Confer- 
ence,* in which Mr. B. T. Lowne observes : " One of the 
difficulties in rearing seedling Orchids arises, I believe, from 
the fact that the pollen is only developed from the prolifica- 
tion of the mother cells, after the pollinia are placed on the 
stigma." He also found that, besides the pistil thus stimu- 
lating the pollen, "the stimulation due to the presence of 
the pollinia gives rise to the development of the capsule, 
even whilst the ovules remain unimpregnated." f 

The significance of the above details lies in the fact that 
external influences, both mechanical and physiological, can 
bring about changes in the epidermal X and sub-epidermal 
layers, with a determination of a flow of fluids of a specific 
character to those specialized tissues. As this is proved to 
be true for the conducting tissues, so do I infer it to be 
equally so for glands of various kinds. 

* Journ. of Roy. Hart Soc, vol. vii. ; see paper by Mr. H. J. Veitch, 
p. 22. 

t L.c, p. 48. " Degeneracy " will be discussed in Chaps. XXYI. and 

X M. Mer found that stomata were developed in the epidermis of 
galls on vine-leaves which normally had none. "Insolation" or 
exposure to light has a marked influence on the orm of the epidermal 
cells, and in increasing the number of stomata. The walls become 
straighter and thicker, and especially the cuticle. M. Mer believes the 
production of stomata to be the direct result of the accumulation of 
nutrient substances. Comp. Rend, xcv., 1882, p. 395. See also Journ. 
Roy. Micr. Soc, 1882, p. 530, and 1883, p. 91. See above, p. 154. 
Another important paper on the same subject, fully corroborating 
these observations, has lately appeared, by M, li. Dufour, entitled, 
Influence de la Lu)niere sur la Forme et la Structure des Feuilles, 
Ann. des Sci. Nat., 7 ser., torn. 5 (1887), p. 311. 




The Laws of Colour. — M. de Candolle proposed to divide 
the colours of flowers into two series, the Xanthic and Cyanic, 
the former containing yellow-green, yellow, yellow- orange, 
orange and orange-red ; the latter, blue-green, blue, blue- 
violet, violet, and violet-red ; red being intermediate between 
the two series. It was thought that flowers were rigidly 
bound by these series, and never transgressed them, but that 
the tints of a species might vary through each. Thus the 
editor of the Gardener s Chronicle, replying to a correspondent 
on Feb. 2, 1842 (p. 97), remarks that "a blue Dahlia was 
not to be expected. On the other hand, the Hyacinth, being 
of the cyanic series, a yellow Hyacinth will not occur." 

Yellow Hyacinths are, however, common enough now. 
Even in 1856, Dr. Lindley found it necessary to conclude a 
leading article on the subject with the words : "At all events 
the cyanic and xanthic speculations of philosophers must now 
be laid up in the limbo of pleasant dreams." 

The many exceptions to this supposed rule met with 
between 1845 and ]856 elicited the above remark, and 
notably a species of Delphinium, viz. D. Cardinale, containing 
" golden yellow in the petals, which are as scarlet as a 
soldier's jacket everywhere else, one of the last of Messrs. 


Veitcli's fine Californian introductions. In this flower there 
is no sign of blue. Yet, if there is a genus more pre- 
eminently blue tlian any other cj'anic race, it is surely 

It is true that some species have never yet transgressed 
their bounds, so that Dahlias still refuse to be bkie now as in 
1845 ; and we are still ignorant of the reason. 

The effect of nutrition upon the colours of plants is well 
known, in that they vary much more in a garden soil than 
in the wild state ; and differ in colouring according to the 
character and ingredients of the soil. Thus, as described by 
a writer in Hovey's Magazine of Horticulture* striped Dahlias 
will be best kept clean by planting them in a poor soil, while 
rich soil invariably runs them. E.g. JD. var. striata formosis- 
sima : No. 1 was planted in a poor gravelly soil, in an open 
situation ; all the flowers but two were beautifully mottled. 
No. 2 was planted upon a rich, cool, sandy loam ; not one- 
half of the flowers w^ere mottled. No. 3 consisted of three 
plants, very highly enriched ; every bloom but one was self- 
coloured. Similar effects follow on the variegated foliage of 
Pelargoniums, according as they are grown in a too rich soil 
or light one.f 

" Alum is said to render the Hydrangea blue ; and some 
saline substances, such as phosphate of iron and muriate of 
ammonia, appear to brighten the tint of red." J It often 
happens, however, that blue and pink corymbs occur on the 
same plant of Hydrangea. A cutting taken from a blue 
Hydrangea growing at Southampton, and transferred to 
Bedfont, changed to the usual colour on blooming there. § 

Chloride of lime has been known to make a whole-coloured 

* Quoted in the Gard. Chron., 1842, p. 8. 

t Gard. Chron., 1876, p. 567. % Ihid., 1843, p. 577. 

§ Ihid., 1886, vol. xxvi., p. 118. 


Camellia become striped; while ammonia enhances the 
colours of Balsams. 

Oxidization is believed to have great inflaence in chang- 
ing the colours of plants, just as it affects certain juices when 
exposed to the air. Thus, if a leaf of the Socotrine Aloe be 
injured, the juice is at first violet in tint, but it soon turns 
to brown. If a potato be grated, the pulp rapidly browns in a 
similar way. Many fungi, especially noted for their poison- 
ous properties, turn blue on injury, as species of Boletus. 
Moreover, they do not do so if exposed to nitrogen, hydrogen, 
or carbonic acid; hence it is presumably the oxygen which 
effects the change. 

Some flowers change their colours from their first open- 
ing to a full expansion : such as Cobcea, from green to violet; 
several Boraginaceous plants, from red or even yellow to blue- 
purple. Lycium harbarum, the popularly called " Tea-plant,'* 
is a well-known instance. Others change during the day, as 
the " Changeable Hibiscus.'^ This plant has flowers white 
in the morning, pink at noon, and bright red by sundown.* 
Similarly, a Phlox of a bright pink colour, " in the early morn- 
ing, by five o'clock, has its colour of a lightish blue, which 
continues to alter as the sun advances, and by nine or ten 
o'clock becomes its proper colour ; the clump which catches 
the sun's rays first changes first, while the other is still blue." 

Though referring these and other well-known instances to 
oxidization. Dr. Lindley, from whose leading article the above 
remarks are partly taken, concludes with the observation, 
" In fact, we know very little about the cause of changes in 
colour, either in plants or animals." Perhaps it remains so 

The intensity of the colours of many high Alpine flowers 

* According to M. Ramon de la Sagra; quoted in Gard. Chron., 1842, 
p. 555. 


bas often been noticed ; and when plants growing near Paris 
were transferred to a much higher latitude, the flowers 
deepened in colour. This, however, is thought not to be due 
to a clearer atmosphere, but to the enhancement of the foliage, 
as M. Ch. Flahault showed that the leaves of plants of the 
same species are larger in proportion as the latitude is 
higher, the comparatively large dimensions being due to the 
duration of light of a relatively feeble intensity. Flowers 
being dependent upon leaves, great importance must be 
attached to the power of the latter to store up nutriment for 
them. Thus, in the case of Hj^acinths both blue and red, 
M. Flahault found no difference in the colour of the flowers 
when grown in the light or in the dark, the colour being at 
the expense of the material stored up in the bulbs. Other 
experimenters have found that, while some flowers show no 
difference, others do ; thus Askenazy found no difference in 
Tulips and Crocuses, though the leaves were etiolated in the 
dark. With Hyacinths, however, contrary to Flahault, he 
found that light exerted a two-fold influence, an acceleration 
of at least a fortnight in flowering and a much more intense 
and more diffused colour ; those in darkness being only 
tinged where the uncoloured ones where darkest. Pulmonaria 
officinalis in darkness changed from red to blue, as usual ; 
but in proportion as the buds were in a less advanced stage 
when placed in darkness, so were the colours fainter. His 
conclusion is that some flowers require light to develop their 
normal colours, while others are independent of it.* Mr. 
Sorby t agrees with Askenazy ; and concludes that the arrest 
of normal development in darkness varies with the nature 
of the colouring matters, the effect being greater with the 
more easily decomposable substances. " Those substances 
which when dissolved out from the petals are the most easily 
* Bot. Zeit, Jan., April, 1876. t Nature, April 13, 1876. 


decolorised by exposure to light, are formed in relatively 
greater amount when the flowers are grown in the dark. 
This is easily explained if we assume that a higher vital 
power, depending on the presence of light, is necessary to 
overcome the more jDOwerful chemical affinities of the less 
stable compounds." 

The crossing of flowers is w^ell known, and much practised 
by florists, to enhance the variety of tints. The interpreta- 
tion is that crossing is a stimulating process, and provokes 
the petaline energy to a high degree. 

From the preceding remarks it will be now gathered 
that colours, per se, are a result of nutrition; and that the 
prevalence of brighter colours in conspicuous flowers which 
are regularly visited by insects is due to the stimulating 
effects which they have produced, thereby causing more 
nutritive fluids to pour into the attractive organs. 

Besides, however, this general result of brilliant colouring 
there are those peculiar and special displays of bright tints 
distributed in spots and streaks in certain and definite places 
only. These have been called " guides " and " path-finders," 
as they invariably lead to the nectaries. If the theory be 
true which I am endeavouring to maintain throughout this 
book, all these effects are simply the direct results of the 
insects themselves. The guides, like obstructing tangles of 
hair and nectaries, are always exactly where the irritation 
would be set up ; and I take them to be one result of a more 
localized flow of nutriment to the positions in question. 

Instead, therefore, of a flower having first painted a petal 
with a golden streak to invite the insect, and to show it the 
right way of entering, the first insect visitors themselves 
induced the flower to do it, and so benefited all future 

The Origin of Colours. — Mr. Grant Allen has written 


an interesting little book on The Colours of Flowers,'^ in 
which he expresses his belief that the first colour on depart- 
ing from the primitive green was yellow. When Ave remem- 
ber that the spore-cases and spores of Lycopodium, tbe 
anther-cells of Cupressus, and the whole anther-scale of 
Pinas and all the pollens of Gymnosperms are yellow, — again, 
when we come to Dicotyledons and find the prevailing tint 
of stamens is the same, — we gather probabilities in support 
of that view. That Nature next introduced reds, and only 
lately, so to say, succeeded in manufacturing blues, seems 
probable from the comparative rarity of the last colour. 
Moreover, when flowers individually change from one colour 
to another as they develop from the bud to maturity, it is 
always in that order — i.e. from reds to mauves or purples, as 
in Echium, Fuhnonaria, etc., or even from yellows through 
reds to purples, as in Myosotis versicolor, so that we still seem 
to gather additional support to the theory. 

If, however, we ask what has caused these changes, we 
are as yet in the dark. A few hints are attainable, and that 
is all. Yellows and reds seem to be due to substances allied 
to the oxidized products of chlorophyll in autumn leaves. 
Again, chlorophyll grains on turning yellow in fruits (Lyciiim) 
become angular, two or three pointed, and finally granular. 
Tn the same way tbe yellow granules of petals {Cucurhiia) 
resemble " amyloplasts," or starch-forming corpuscles. f 

The general conclusion one arrives at from various 
observations is that the original change from the ancestral 
green to, probably, yellow is correlated to the change of 
function; but why the first colour was yellow, and why it 
ever gave place to red or blue, is unknown. 

Supposing the yellow-green colour to have spread to the 
adjacent parts which then attracted insects, as it does in 
* la Nature series. f Sachs' Veg. Phys., p. 320. 


some 'Eupliorhias and Chrysosplenium, for instance, then the 
visits of insects Avould bring the required stimulus to advance 
tlie colour to a pronounced yellow ; and so petals, it may be 
conceived, came into existence. 

Pale and White Varieties. — The paler tints or even a 
total absence of colour may seemingly occur as a variety of 
any plant. It is often a concomitant of habitual self- 
fertilisation in cases where the variety or species is a 
degradation from some conspicuous and biightly coloured 
insect-visited form. White-flowered individuals often appear 
as " sports " amongst seedlings ; the immediate cause of 
which it would be difficult to assign, beyond the general 
one of the absence of those nutritive conditions which are 
requisite for cjlours, as occurs in Gladioli* 

White, however, is useful as a starting-point for florists' 
flowers where great variegation is required. Thus M. 
Vilmorin | says that " in ten examples of variegation 
which were produced under my own observation, the course 
was always the same. The original plant, with flowers whole- 
coloured, gave in the first instance a variety of flowers 
entirely white ; afterwards, variegations were produced from 
this white variety on its returning towards the coloured type. 
. . . This pure white variety usually gives in the first sowing 
a greater or less proportion of plants with flowers like those 
of the coloured type ; but by careful selection through 
several generations the pure white type is in most cases 
completely fixed. ... It is only among the white varieties not 
completely fixed that the variegations make their appear- 
ance ; at first they exliibit narrow pencillings, the coloured 
portion being only one-tenth, and sometimes only one- 
twentieth of the whole surface ; but then in the following 

* Garden, 1880, p. 327. 

t Flore des Ser. et des Jard. de VEur., {Gard. Chrun., 1852, p. 500). 


generation . . . the coloured portions begin to predominate. 
... I have never been able to observe a single instance of 
variegation coming directly from the coloured original. The 
contrary, however, takes place with regard to the dottings ; 
these come directly from the coloured tjpe." The variegated 
varieties the author had succeeded in fixing were Gomphrena 
glohosa ; Antirrhinun Tnajus, Convolvulus tricolor, Neifnopliila 
insignis, Porfulaca grandiflora and BelpMnium Ajacis. 

Other florists have found that by crossing whole-coloured 
flowers pure white seedlings may result. 

Abutilons have an instructive and, in part, a somewhat 
similar history. No hybrids were raised from the old bronze- 
red and striped form, which was usually barren, until the 
white "Boule de Neige " was introduced. Mr. George 
crossed this with " Duke of Malakoif." The white one had 
itself previously thrown up every shade of dingy white ; but 
whether by being spontaneously crossed or uot, does not 
appear to be known. Some of the colours of the seedlings 
of this cross were pale and dark pink, pale orange, bright 
carmine, salmon, orange-red, etc.* 

Somewhat analogous results w^ere obtained by Mr, Veitch 
with Rhododendrons imported from Borneo. Thus a cross 
between the larger-flowered B. Javanicum, which is orange- 
coloured, with the smaller white narrow-lobed B. Jasmini- 
florum, gave rise to the rose-coloured " Princess Royal." A 
further cross of the last with the parent B. Jasminiflorum 
eliminated the red colour ; the offspring, however, retained 
the form and size of the corolla of the " Princess Royal." 
It is called " Princess Alexandra." 

In the above-mentioned the effect of the white has been 
to separate the tints; so that from the old Bronze-red 
Ahutilon JDarwinii we get yellows and reds of different 

* Gard. Chron., 1878, p. 792. 


shades. Similarly the orange or buff-yellow Bhododendron 
Javanicum has been split up into vj.rious reds; the white 
having, so to say, eliminated the yellow. 

The subsequent eii'ect of crossing with regard to flowers 
is variety. With this fact florists and horticulturists are 
familiar : for as soon as crossed or hybrid seedlings are 
raised the varieties of colouring become infinite. It has 
been observed that the " spots " are more persistent than 
the base-colour of the flower. This fact agrees with the 
theory advanced that they have, whenever they occur as 
guides or path-finders, been determined by the insects and 
then become hereditary as much as the shape of the flower 
itself ; and as that is maintained much more persistently 
than general colouring, so is that specialized colouring which 
has been equally due to insect agency. 

With regard to the correlations which exist between 
colours and insect visitors, Miiller especially has observed 
several. Thus beetles seem to affect yellows, e.g. TJialictrum 
and Galium verum ; wasps and carrion insects, reddish- 
browns, such as of Comarum, EpipacHs, etc., while the more 
intelligent bees, etc., delight in purples and blues ; and it is 
thought that their selective agency has determined the sur- 
vival of such special colours as they prefer. This has been 
probably the case, but we still want to know what is the 
immediate cause which induces one colour to change to 

As high colouring or conspicuousness if the flower be 
white is due to insects, so pale colouring and inconspicuous- 
ness is due to their absence ; but what the nature of the 
stimulus is we cannot tell. It enhances the assimilative 
powers ; for the crossed plants, as Mr. Darwin abundantly 
proved, are usually larger plants. It usually infuses some 
of the characters, floral or foliar, of the male parent — but 


not always : several experimenters assert that, after every 
precaution, the offspring exactly resemble the maternal 
parent. But one rule florists always adopt in order to 
enhance the colouring is to use the pollen of the better- 
coloured plant, the maternal parent being usually the in- 
ferior one. 

As an illustration of the relative effect, of crossing and 
self-fertilisation respectively on the production of colours, 
I quote the following passage from Mr. Darwin's work : * 
" The flowers produced by self-fertilised plants of the fourth 
generation [of Dianthus caryopliyllus or Carnation] were as 
uniform in tint as those of a wild species, being of a pale 
pink or rose-colour. Analogous cases [occurred] with 
Mimulus and Ipomoea. . . . On the other hand, the flowers 
of plants raised from a cross with the fresh stock which bore 
dark crimson flowers, varied extremely in colour. . . . The 
great majority had their petals longitudinally and variously 
striped with two colours." 

Uniformity and paleness of tint are thus correlated Avith 
self-fertilisation; and since, whenever the latter process is 
persevered with, an increase of fertility follows, it is not 
surprising to find that such tints are usually accompanied 
by an increased power of seed-bearing. Thus, Mr. Darwin 
found that, "the proportional number of seeds per capsule 
produced by the plants [of Dianthus'] of crossed origin, to 
those by the plants of self -fertilised origin, was as 100 : 125." 
Again, of Antirrliinum majtis, the relative self-fertility of red 
and white varieties was as 98 : 20 ; of Mimulus luteus the 
same comparison gave the ratio of 100 : 147; while pale- 
coloured Pelargoniums are notoriously great seeders. "f 

* Cross and Self Fertilisation, etc., p. 139. 

f For further illustrations, see mj paper on Self.fertilisation, etc. 




Theoretically, as already stated, a perfect flower sLould or 
might be composed of six whorls, if its parts be not spirally 
disposed, — the perianth, andrceciam, and gyncecium. each 
consisting of two verticils. The very general rule for their 
emergence from the axis is centripetal. The subsequent 
rates of development of the several whorls may vary con- 
siderably, so that one part which emerged first, or at least 
very early, may be late or the last to arrive at maturity. 

The calyx or outermost whorl of the perianth when 
present is nearly always the first to appear, and to grow 
rapidly to a relatively large size, and thus protects the more 
rudimentary parts within it ; but if it ultimately remains 
rudimentary itself, or, it may be, is not entirely arrested, then 
it is the corolla which first emerges, the function of pro- 
tecting the essential organs being relegated to it. Such is 
the case with the Compositcp, Valerianece, etc. 

The cor-oUa, with rare exception, emerges before the 
stamens, though it is very generally rapidly passed in 
development by the latter organs. In Lopezia and Primula, 
however, the stamens emerge first ; and this has led some 
botanists* to regard the petals of the last-named plant as 

* For references and literature on the stmctnre of Primulacece, see 
Masters's paper, On some Points in the Morphology of the PrimulaceiJBf 
Trans. Lin. See, 2nd series, Botany, vol. i., p. 285. 


outgrowths from tlie stamens. My oAvn observations tend 
to coii6rm those of Dr. Masters, that it is an exceptional fact, 
and not constant. It appeared to him " that in Lysimachia 
Nummularia the petals did really sometimes (but not always) 
precede the stamens in their development." 

The stamens emerge before the pistil, and if there be two 
whorls to the androecium, it is the sepaline whorl which 
appears first ; though the fully developed stamens sometimes 
assume a position, as already explained, within the petaline, 
as in Geraniacece. Like the corolla and staminal whorls, the 
carpellary appears all at once, and last of all. 

With reference to the emergence of the individual parts 
of the whorls, it is an almost invariable rule that those of the 
outermost whorl of the perianth or calyx, if it consist of three 
or five parts, rise centripetally in succession according to the 
laws of phyllotaxis. Thus, if the calyx be pentamerous, its 
parts invariably emerge in quincuncial order, thus consti- 
tuting a cycle of the f type. . If it be trimerous, as in Mono- 
cotyledons, it is a cycle of the J type. If, however, it be 
tetramerous, then the parts emerge in decussating pairs, as 
in Tamarix tetrandra, Sparmamiia, Philadeljphus, and the 
sepals in the Cruciferce* This clearly shows that a normally 
tetramerous calyx is the result of the combination of two 
pairs of leaves, corresponding to two nodes, the internode 
between the pairs being suppressed. 

The parts of the inner whorl of the perianth or petals of 
the corolla, as also those of each staminal and carpellary 
whorl, almost invariably emerge simultaneously if the whorls 
be regular ; though pronounced differences may occur in the 
case of in'egular flowers. Similarly, when there is a strong 
spiral tendency, as in the RanunculacecB, members may arise 

* The lateral sepals, though overlapped by the other pair, are the 
arst to receive their vascular cords from the axis. 



successively. If tlie stamens be very numerous tliey 
usually emerge in centripetal order, as in Buttercups ; but 
tliey may form " centrifugal groups," as in Hypericum ; the 
numerous stamens of Cistus and Helianthemum, as well as 
of Cactus^ Opuntia, and Mesevihryanthemum, and the Loasece.^ 
are also centrifugal in their development. Lastly, if the 
carpels form a wliorl, they, too, emerge simultaneously ; but 
if they be numerous and spirally arranged they emerge and 
develop in succession. 

There are some additional points to be observed. The 
first is the method of change from tetramerous to pen- 
i 2 s 

tamerous in the same plant. Thus in Celastrus scandens, 
if the flower be tetramerous, the sepals appear in pairs, the 
antero-posterior first, then the latei'al pair afterwards. If 
the flower be pentamerous the sepals arise in succession 
quincuncially, the numbers 1 and 3 being anterior ; numbers 
4 and 5 are lateral, and number 2 posterior. 

Now, by referring to the diagrams above, it will be 
seen that this order is in exact agreement with the usual 
method of passing from opposite to alternate arrangements 
in the foliage. The correct angular distance or divergence 
being acquired immediately in the case of the calyx, by 
shifting the position of the parts so that the divergence of 
144° is obtained. In the case of foliage, this is only secured 
after several internodes (see p. 18). 


Exactly the same procedure occurs in Sparmamiia 
and Philadelphus, which are tetramerous, as compared with 
Tilia and Ueutzia respectively, which are pentamerous (see 
p. 18). 

The next point to be noticed is the alteration in the 
order of emergence which takes place in irregular flowers. 
The rule seems to be that those parts of the flowers which 
assume a greater prominence in the mature state, or have 
some special function beyond the rest, emerge and develop 
before the others. Thus in Leguminosce and Lahiatce, where 
there is a prominent "landing-place " for insects, the petals 
issue successively in an antero-posterior order. The carina 
of papilionaceous flowers composed of two petals appears 
first, then the al^ together, and finally the vexillum. In 
Reseda, the sepals, petals, and stamens issue in a postero- 
anterior manner ; but while the sepals finally attain to much 
the same dimensions, the petals remain more or less atrophied 
as they emerge towards the anterior side. Then the stamens 
appear in the same order upon a cellular ring, which, later 
on, grows out into the unilateral disk between the petals and 

In a few regular flowers the simultaneity is also wanting . 
thus in Adoxa the sepals of the tetramerous terminal flower 
emerge in pairs, and the four petals simultaneously ; but in 
the lateral flowers the posterior sepals issue before the 
anterior; and of the five petals the posterior one emerges 
first, the two lateral secondly, and the two anterior ones last 
of all. These modifications are continued in the order of 
flowering. Thus the terminal flower expands first, and " all 
at once." Of the lower lateral flowers the two upper 
posterior sepals open out first, then the posterior stamens 
mature and shed their pollen. The anthers dehisce in suc- 
cession from the lateral stamens, and lastly from the anterior 


ones. The lower sepals do not separate until after the upper 
stamens have shed their pollen.* 

Though we are not in a position yet to account for all 
such deviations from general rules, yet I think in such cases 
as the Legiiminosce and Labiatce, and probably all irregular 
flowers, that the rationale may with great probability be 
assumed to be the stimulus given from without to meet the 
extra strain which certain petals or stamens or both have to 
sustain while supporting the weight of an insect when visiting 
them. To meet this demand an extra supply of nutriment is 
sent to the parts which thus require it ; and, in fact, I believe 
the final result has thus been actually brought about by the 
effort of the plant itself, so that it has developed parts in 
accordance with its requirements in a manner parallel with 
that which has obtained in the animal kingdom. 

In the case of Adoxa I would regard the above-mentioned 
orders of development as a result of unequal distribution of 
nutriment in order of time. Thus the apical flower receives 
its nutriment first and develops first ; then the other flowers 
which are placed laterally subsequently. And this order of 
supply has affected the parts of the latter flowers in the 
same way, so that they develop from above downwards, or 
in a postero-anterior manner. It may be compared to a 
three-flow^ered cyme, of which the central flower expands 
first, and the two lower ones afterwards. 

A feature must here be noticed, though I do not think much 
stress need be laid upon it, which botanists have called "obdi- 
plostemony." f If a flower have one whorl of stamens of the 
same number as the petals it is isostemonous ; of two, diploste- 
monous ; and if the stamens of the outer whorl be opposite or 

* For a note on Adoxa, see my paper On the Origin of Floral JEstiva- 
tions, Trans. Lin. Soc, 2nd series, Botany, voL i., p. 194. 
t Sachs' TexUBooTc, 2nd edition, p. 601. 


superposed to the petals, and therefore antipetalous, then the 
above term is used : for the rule is that the calycine whorl 
shoald be outermost and emerge first ; then the petaline, which 
usually takes a position higher up on the axil ; and, in at 
least most of the genera and orders where obdiplostemony has 
been noticed in the completely developed flower, it is simply 
due to the petaline whorl of filaments being, so to say, thrust 
outside the level of the calycine whorl by the protruding 
buttress-like bases of the carpels, as in Geranium pratense. 
This is still more the case in Oxalis, Avhere, as in Geranium, 
the sepaline stamens become the taller set, the petaline the 
shorter; and the position of the former being more internal 
than usual, apparently in consequence of the appendages 
which grow on the outer side of the filaments.* 

Again, the order of emergence may be the same as usual, 
namely the sepaline stamens first, then the petaline ; but 
the position of the latter, instead of being within as is 
the rule, may be apparently on exactly the same plane as 
the sepaline, as in Heaths. Since, however, 
they do not emerge simultaneously, but one 
set is intercalated between the other, or even 
outside of it (Fig. 51), this order of appear- 
ance is, to my mind, a sufiBcient proof that 
they do not really belong to the calycine pig. 51. -Diagram 

1 1 of emergence of 

wiiuii. petaline stamens 

There is no greater difficulty in under- ^fde'^TJepaira; 
standing this, than in seeing that a compres- (afier Payer), 
sion of the internodes of opposite and verticillate leaves has 
taken place when double the usual number are present in 
a whorl. Thus privet has sometimes four leaves at one node, 
forming a quaternary whorl, and all on the same plane ; and 

* According to Frank, in OxaUdece and Geraniacece, it is the anti- 
petalous stamens which are developed first. See above, p'. 150. 


this will remind the reader that, since floral whorls, are based 
upon pbyllotaxis, ten stamens could not possibly form a 
cycle ; and although the eight stamens of a Heath might do 
so, there is nothing in the leaf arrangement of that genus to 
suggest their being a whorl of the | type. 

Since the petaline cords are usually united to the 
staminal ones, the fact that the petaline stamens get 
sometimes, as it were, " dragged outw^ards," offers really 
no great difficulty ; but is, so to say, a mere accident brought 
about by the adaptations of the flower to insect agency. 

Indeed, to interpret these irregularities in the emer- 
gence, one must look to the final condition to see if there 
are any ultimate results in correlation with them. In Oxalis 
we get heterostylism with its corresponding different lengths 
of the filaments, and the necessary adjustments of the latter; 
since there are at least two sets in each flower, for insects to 
readily secure the pollen. In Heaths all the anthers are 
arranged in a ring round the style, pressing their cells 
against it, and so closely approximated, that when a bee 
dislocates one by pushing the lever-like auricle to one side, 
she dislocates the whole, and so receives a shower of pollen. 

These final arrangements, therefore, are suggestive of the 
reason why the points of emergence of the stamens occur just 
where they do. 

In the case of Hypericum, where the stamens emerge 
centrifugaUy, from a definite number of original papillae, 
three or five as the case may be, the stigmas extend 
outwards ; so that, if they have not been pollinated by 
insects, they can come in contact with the latest formed or 
the outermost anthers. 



The order in which the several whorls of flowers emerge 
from the axis is, as stated above, almost invariable ; but the 
rates of development are very various, and important sexual 
and other differences follow as the results. For flowers with 
conspicuous corollas or other structures attractive to insects, 
the prevailing order of progression subsequent to emergence 
is first the calyx, secondly the stamens, and, if tbere be two 
series, the whorl superposed to the sepals grows first, 
afterwards the whorl superposed to the petals ; then follows 
the pistil to a point approaching maturity, when the corolla, 
just before expansion, grows very rapidly to its full size ; and 
finally the stigmas mature. The anthers have also grown 
long before the filaments, which at last elongate very rapidly. 
The usual result on maturity is various degrees of protandry, 
coupled with conspicuousness or attractiveness to insects. 
As a few of the examples I have examined maybe mentioned 
Kanunculus acris, Cardamine pratensis, Stellaria Holostea, 
Lychnis dioica (male), Malva moscJiata, Geranium (larger 
flowered sp.). Pelargonium, Tropceolumj Epilohium hirsutum, 
(Enothera biennis, Ipomcea, Veronica Chamcedrys, etc. In fact, 
this order of growth and development prevails generally with 
flowers having conspicuous corollas. 

The interpretation appears to be as follows. In such 


flowers as these, energy is especially directed into the 
development of the corolla and androecium ; the former 
being large, and the latter supplied with much and often 
highly differentiated pollen. All this means the consumption 
of so much more nutriment ; and, as the chief amount of 
floral energy is thus directed first into the androecium, then 
into the corolla — which often attains a far greater size than the 
other organs, — consequently these tw,o whorls tend to draw 
a large amount of nourishment to themselves. In conse- 
quence of this, the pistil has, temporarily at least, to 
suffer ; so that its growth is for a time delayed, and it 
does not mature as early as the stamens, which had, moreover, 
a considerable start in the race to maturity. Hence the 
result is that the stamens are often mature and even shed all 
their pollen long before the stigoaas are prepared to receive it. 

This, then, accounts for protandry being almost invariably 
the rule in the case of relatively conspicuous flowers.* 

If flowers have two or more whorls or many series of 
stamens, as have many genera of Canjophyllece, GeraniacecB^ 
BananculacecB, and RosacecB, then the pistil may arrive at 
maturity between the periods of different series, or con- 
temporaneously with some of them ; so that, while the flower 
is protandrous with regard to the first stamens which mature, 
it is homogamous with others, and thus self-fertilisation can 
be readily secured if the flower fail to be crossed. 

It may be here observed, though the fact will be dwelt 
upon again, that by far the greater majority of flowers, 
conspicuous or not, retain this provision for self-fertilisation ; 
and that those flowers which normally cannot possibly 
fertilise themselves are in a very small minority. 

* There are a few protogynons flowers, it is true, which are more 
or less conspicuous, but these exceptional cases have their own inter- 
pretations, which will be considered later on (see Chap. XXII.). 


Nearly the same order of development as the above is 
maintained with some that have rather inconspicuous flowers 
in consequence of the corolla being small ; but then it must 
be remembered that the other organs are proportionally- 
small too, and, if they come at all, are visited by small 
insects. Such, for example, are Malva crispa, Veronica 
serpyllifolia^ V. agrestis, etc. In these flowers, however, the 
pistil has a remarkably rapid growth as compared with the 
preceding cases. The cause is, that energy is now directed 
at once to that organ, instead of being so largely occupied by 
the- stamens and corolla. The result is that the pistil 
matures more rapidly than in the previous cases, and 
sometimes even simultaneously with the stamens. The 
flower is therefore more nearly homogamous, and self- 
fertilisation can with them more easily ensue. 

In many cases amongst inconspicuous flowers I could 
detect no appreciable difference at all in the rates of 
development of the essential organs. I would then describe 
the order as Calyx, Stamens -f Pistil, Corolla. As examples 
are Lejpidium campestre, Sisymhrium AlUaria, and S. officinalis, 
Nasturtium officinale, Corrigiola littoralis, (Enothera historta, 
etc. These are all, it will be noticed, very small-flowered 
plants. They are thus homogamous, and habitually self- 

The next order of development to be noticed is Calyx, 
Stamens, Corolla, Pistil. As far as my observations go, this 
order appears to be mainly confined to gamopetalous flowers, 
with a hypogynous corolla, as Linaria minor, L. Cymhalaria, 
Veronica spicata, Primula,* Ancliusa officinalis, Borago offici- 

* This order of development in Primrose has been observed by- 
others, and apparently thought to be exceptional ; so that the somewhat 
strange suggestion of the corolla being an outgrowth of the andrcecium 
was made by Pfeffer ; but it by no means stands alone in this respect. 
See Sachs, I.e., p. 609 ; Jahrh.fiir Wissensch., Box., vol. vii., p. 194. 


nalis, Amsinckia angiistifolia, Statice psilocladia and Plantago 
Coronopus, etc. 

The remarkable delay in the progress of the development 
of the corolla during the emergence and first stages of 
development of the stamens is the peculiar feature. It 
sometimes allows the stamens to emerge first, as in Primula ; 
or if they be nearly simultaneous, then the corolla may be 
suddenly checked, as in Veronica. But many differences 
occur ; thus they emerge and grow up together in Samolus, 
while in Ancliusa officinalis the corolla rapidly exceeds both 
stamens and pistil. In the case of Amsinckia the corolla 
and stamens appear to emerge almost together, and then 
follows the pistil, which the former quickly exceed in height. 
Then the pistil regains the height of the stamens, and they 
ultimately mature together. A similar procedure obtains 
with Plantago Coronopus : though the petals emerge first, 
the anthers quickly outstrip them, and the corolla grows 
considerably more than the pistil, which is consequently 
delayed ; but when they are nearly developed and the corolla 
becomes scarious, then the style elongates with great rapidity, 
and the stigmas mature first, so that the flower is ultimately 
protogjnous. Exactly the same course is followed by the 
floral whorls of Statice psilocladia. 

The next order of development is Calyx (if present), 
Corolla, Stamens, Pistil ; or even Corolla, Calyx, Stamens, 
Pistil. The cause of the corolla developing so soon is the 
arrest of the calyx, as in Umhelliferm, Valerianece, and Com- 
positce. The corolla now has to act as a protecting organ, and 
always keeps in advance of the essential organs. Indeed, in 
the orders with epigynous and gamopetalous corollas, in which 
the calyx is usually obsolete or nearly so, the corolla actually 
emerges before it. 

The last order of development to be mentioned in the case 


of flowers possessing a corolla is Calyx, Pistil, Staniens, 
Corolla. As examples, I find the following illustrate this 
condition : Hanunculus sceleratus, Cardamine liirsuta, Cerastium 
glomeratum, Arenaria trinerva, Sagina pi'ocuinhens, Spergu- 
laria marina, Polycarpon tetraphyllum, Trifolium minus, 
JEpilobium mo7ita7ium, Gaura parviflora, etc. This appears to 
be the most general condition for very small and incon- 
spicuous flowers which are regularly self-fertilised. The 
interpretation is the exact converse of the order of develop, 
ment first described ; namely, of the whorls of conspicuous 
flowers. All the above are inconspicuous, many being rarely 
if ever visited by insects ; and as the corolla is minute, no 
nourishment is required for the petals, the stamens are often 
reduced in number and the quantity of pollen diminished. 
The pistil at once proceeds to grow, and the result is, if not 
homogany, protogyny. 

It must be now borne in mind that the above differences 
in the order of growth and development must not be regarded 
as at all absolute or invariable, but only general rules as to 
what takes place ; for the rates of growth of the respective 
whorls may vary in the same species according to external 
circumstances ; so that a plant may be protandrous at one 
time or place, homogamous or even protogynous elsewhere 
or in another season, as the case may be. Indeed, Miiller 
frequently calls attention to this fact, to which I shall have 
occasion to return. 

Emergence and Development of the Ovules. — If the 
ovules be tolerably numerous, the order in which they appear 
is not constant. It may be either from above downwards or 
from below upwards on the placenta. Thus, as Payer has 
shown by his drawings, in Viola, Reseda, Cistus, Tetrapoma, 
Fiimaria, Linicm, Rttta, Meliantlius, Staphylea, Spircea, and 
Opuntia the order is basifugal, or from below upwards. On 


the other hand, in Macleya, Dicentra^ Epimediiim, Bartonia, 
ImpatienSf Lythrum, Dracojphylluvi, 3IalacMum, Cerastiuin, 
Fri7)i2ila, and Samohis the order is basipetal, or from above 

When the row of ovules is very numerous, it is the rule 
that the point where thej first begin to emerge is midway, 
and the development takes place both upwards and down- 
wards simultaneously. It is thus with Selleborus and allied 
genera with follicles, Capparis, Epilobium, Trifolmm, Cajo- 
plioraj Latliyrus, Citrus, Fassifiora, and the Monocotyledonous 
orders, Iridacece and Amaryllidacece. Lythrum, and Opuntia, 
however, both of which have considerable rows of ovules, 
develop them, as stated above, in a basipetal and basifugal 
manner respectively. 

On examining Payer's numerous figures, I find that when 
the order of development is from below upwards, the OTules 
have their micropyles upward ; when they develop from 
above downwards, the micropyles grow downwards. In 
either case, occasionally the middle ones may be somewhat 
horizontal, if they are somewhat numerous, as in Bartonia, 
Spircea, and Sfapliylea. When they are very numerous and 
develop both ways from a point midway, then the ovules 
may either turn upwards or downwards ; the majority being 
downwards in the proportion of nine to five. 

As a theoretical interpretation to account for the general 
fact of the central ovules developing first when there are 
long rows of them, it may be due to the carpel being com- 
parable to a lanceolate leaf, where the longest and therefore 
the most vigorous nerve-branch of the pinnate nerves is in the 
middle. If the rows of ovules emerge from below upwards, 
the carpel may be comparable to a more primitive type, as of 
monocotyledons with a palmate foliage. Thus the only 
exceptions I can find in Payer's figures of Monocotyledons are 


the Gladiolus and Alstroemeria, where they are very numerous 
and follow the rule of commencing to emerge in the middle, 
and then proceed upwards and downwards. Though parietal 
placentas seem generally to have their ovules developed from 
below upwards, yet, as seen above, it is not uncommon with 
an axile placentation. If any interpretation be sought, I 
should feel inclined to associate it somewhat Avith a more 
primitive state of things, since a parietal placentation presents 
a more rudimentary character than an axile. But icliy they 
are developed thus, sometimes upwards, sometimes down- 
wards, or both ways at once, is at present as inexplicable as 
the fact that leaves develop both basipetally and basifu gaily, 
either in their entirety, or as to their lobes and notches, which 
may be formed on either plan. Perhaps there may prove to 
be a common cause for both. 



Heterogamy* and Autogamy. 

Protandry, Protogyny, Homogamy, and Cleistogamy. — These 
conditions prevail in nature in varying degrees of frequency. 
The first is common to all conspicuous flowers habitually 
visited by insects, and is accompanied by heterogamy. The 
fact that anthers mature their pollen before the stigmas of 
the same flower are ready to receive it, is due to the extra 
stimulus given to the andrcecium, which mostly effects 
simultaneously the enhancement of the corolla or perianth 
■which attracts the insects (see p. 191). Like everything else 
in nature, it is very far from being absolute, and any flower 
may be protandrous at one time or place, while it may at 
another mature the essential organs together, and then it 
becomes homogamous, or it may be even protogynous. 

These latter conditions prevail in less conspicuous flowers 
and all those which are fluctuating between a condition 

* Heterogamy, i.e. union by intercrossing different flowers. 

Atitogamy, i.e. union by self-fertilising one and the same flower. 

Protandry, i.e. stamens maturing the pollen before the stigmas of 
one and the same flower are ready to receive it. 

Protogyny, i.e. pistil maturing the stigmas before the pollen of one 
and the same flower is shed. 

Homogamy, i.e. pollen and stigmas of one and the same flower 
maturing simultaneously. 

Cleistogamy, i.e. autogamous within an unopened perianth. 


requiring insect agency and self- fertilisation or autogamy; 
as well as in the majority of flowers whicli are too incon- 
spicuous to invite insects at all, or which never expand. The 
series of such flowers terminates in perfect and perpetual 

The first condition, or Protandry, does not now require 
special discussion or illustration ; as it is the prevailing one 
iu most conspicuous flowers : though it must be distinctly 
borne in mind that the exceptions are rare in which a flower 
cannot fertilise itself at some period or other before it fades ; 
even though a large order, as Orcliidece, may furnish many 

Protogyny may arise from several causes. Miiller has 
mentioned about twenty species of plants irrespective of the 
Grasses whicli are more or less decidedly protogynous ; and 
what one notices is that many are Alpine species of genera 
which have other species dispersed elsewhere that are homo- 
gamous or protandrous. Thas A7iemone aljpina is protogynous, 
but A. Narcissifolia is protandrous. Banunculus montanus, It. 
parnassifolius, R. pyrenceus are all protogynous. These may 
be compared with the smaller-flowered forms of R. aquatilis 
which are homogamous ; but B. flammida^ R. acris, R. repens 
and R. hulhosus are protandrous with the outermost stamens 
only. Thus, this genus supplies a progressive series. Other 
protogynous and mountain species are JDryas octopetala, 
species of Saxifrage, as S. androsacea and S. viuscoides, and 
^S'. Seguieri : but Miiller found S. oppositifoUa and S. tridac- 
tylites to be sometimes feebly protandrous, at others proto- 
gynous. On the other hand, S. rotundifolia, 8. aizoides, etc. 
are protandrous. Loiseleuria procumhens, Trientalis Europcea, 
Bartsia alpina, Hutchinsia alpina, and TJialictrum alpinum 
are all protogynous. 

Secondly, a group of plants, the flowers of which have 


the habit of blossoming early, as in the spring or the begin- 
ning of the summer, are protogjnous ; such are species of 
Hellebore, Prunus, and Cratcegus, as well as the Horse-chestnut 
and Mandragora vernalis. 

Some species are characterized by the habit of living in 
shady places, as Geum urbanum and G. rivale, Ghrysosjplenium 
ojppositifolmm, Gagea lutea, Paris quadrifolia. 

Lastly, others have minute flowers, as Geranium jpusillum, 
Veronica serpyllifoUa, Toffieldia, and many other species, some 
of which I have mentioned when treating of the emergence 
and developmeut of the floral whorls, where I have explained 
the cause. * 

Wind-fertilised or anemophilous flowers are for the most 
part protogynous ; for these flowers have been accompanied by 
strong degeneracy of the corolla and pollen, while all traces 
of nectariferous structures are almost invariably and entirely 
suppressed.! Hence Thalictrum minus, Poierium. Sanguisorha^ 
Plantago sp., Gallitriclie, Myriophyllum, Artemisia, Glieno- 
podiumy AmentifercB, Juncacece, and Graminece are all more 
or less characterized by being protogynous while they are 
anemophilous as well. 

If we are not in a position to trace the actual causes of 
protogyny in every instance, we can at least see several 
influences which can bring it about. Temperature will be 
seen hereafter to be a most potent one ; for a relatively lower 
temperature very frequently checks the energy of the corolla 
and stamens, without having any necessarily corresponding 
effect on the pistil, and several compensating processes then 
come into play ; so, conversely, the pistil now gains the 
ascendancy and can mature first. This, therefore, will 

* See Chaps. XX. and XXI. 

t Intercrossing by insects may be recovered in anemophilous flowers; 
when honey may be again secreted, as in Salix caprcea and Sanguisorha 
officinalis ; see Fertilisation, etc., p. 236, fig. 77. 


account for some mountain species, as well as those blossoming 
early or in shady places, being protogynous. 

It must not be regarded as universally true. If flowers so 
situated or circumstanced be abundantly visited by insects, 
they will respond to their influence ; and the consequence is, 
that many Alpine plants are even strongly protandrous, as 
well as spring-flowering plants and some which grow in 
shady places, as Sanicula Eu7'ojpcea, Odontites serotina, etc. It 
is when we compare the protogynous species with others of 
the same genus, that the influences of a lower temperature, 
shade, etc., more especially suggest themselves as true causes 
of protogyny in some species, while others may be homo- 
gamous or protandrous. 

Many plants normally provided with conspicuous flowers, 
bat accidentally growing in shady places, may often be found 
having them half opened or as quite closed buds, and yet 
fully fertile. The same occurs late in the season, when the 
flowering period is drawing to a close. Such flowers repre- 
sent the preliminary stages leading to a permanently homo- 
gamous or protogynous condition, as the case may be, which 
are mostly autogamous as well. 

Whatever may be the direct cause, and there may be 
others besides those I have mentioned, protogyny is easily 
brought about temporarily in individuals, or it may become 
hereditary and a permanent feature. 

It need now hardly be added that, before protogyny is 
reached and emphasized, all degrees of passage can be met 
with from strong to weak protandry; then homogamy is 
acquired: and, after passing through oscillating conditions, 
permanent protogyny can be finally the result. 

Many individual plants vary in this respect, being some- 
times or in some places in one condition, and at other times 
aind in other places in another condition. As nothing is 


absolute in nature, so in this case, plants respond to the 
influences brought to bear upon them, and each individual 
may vary accordingly, but if the influence be permanent, then 
the variation becomes hereditary, and one or other character is 
fixed, and may be regarded as specific or generic as the case 
may be. Should the environment change again, what may 
have been constant for generations will be once more broken 
up, and instability ensues. 

Miiller records several cases of such oscillations, as in 
Tulsatilla vernalis, Drijas odopetala, Ribes petrceum, Gentiana 
campestris, Veronica serpyllifoUa, V. spicata, Walnut, Hazel, 
etc. These vary from protandry through homogamy to 
protogyny. He also mentions species which have not yet 
arrived at complete protogyny, such as Sibhaldia procurtibens 
and Ranunculus alpestris, mountain species which are homo- 
gamous ; while R. glacialis is sometimes even slightly pro- 
tandrous. Papaver alpinum, Arahis alpina, and Biscufella 
IcBvigata are also described as homogamous. 

As the transitions from a conspicuous, protandrous, and 
entomophilous or insect-fertilised flower to a homogamous 
and autogamous or self-fertilised one, as well as to anemo- 
phily, are the effects of degeneracy, they will be considered 
more fully when that peculiar condition of floral structure 
comes to be discussed.* 

* See Chaps. XXVI. and XXVII. 



Dimorphic Flowers. — A large portion of Mr. Darwin's work 
on the " Forms of Flowers " deals with the varieties and phe- 
nomena of heterostylism, which is specially characteristic of 
the Primulacem, and Rubiacece, though several instances exist 
in other orders as well. He and Mr. J. Scott were mainly 
interested in showing that " illegitimate " or homomorphic 
unions were less prolific than " legitimate " or heteromorphic ; 
and inferentially took occasion to describe the differential 
sexual characters of the forms of the same species. With 
regard to this latter fact, when Mr. Darwin experimented 
with wild Cowslips, he first thought that they were tending 
towards a dioecious condition, and that the long-styled plants 
were more feminine in nature, and would produce more seed : 
conversely, that the short-styled plants were more masculine. 
Contrary to his anticipation, of plants marked growing 
in his garden, in an open field, and in a shady wood, the 
short-styled forms gave most seed, the weight of seed being 
in the proportion of 41 to 34 ; that is, the short-styled pro- 
duced more seed than the long-styled in the proportion of 
nearly 4 to 3. Similarly when a number of wild plants w^ere 

* Heterostyled, i.e. plants with stamens and styles of different but 
corresponding lengths on separate plants. 

Homostyled, i.e. when stamens and styles are of the same length. 

Homo-, di., tri; poly., and hetero-morphic, i.e. flowers of the same, 
two, three, many, and different forms, respectively. 


transferred to his garden, the result was as 430 to 332, the 
weight of seed being therefore nearly 4 to 3. Lastly, of 
plants covered by a net, six short-styled plants bore about 
50 seeds, while 18 long-styled plants bore none at all. 

From these results, Mr. Darwin wrote, " we may safely 
conclude that the short-styled form is more productive than 
the long-styled form. . . . Consequently my anticipation that 
the [long- styled form] would prove to be more feminine in 
nature, is exactly the reverse of the truth." * We shall 
see, however, that his surmise was probably, to some extent, 
right, nevertheless. 

Mr. Darwin and Mr. Scott have recorded a great number 
of experiments in crossing heterostyled plants, and the 
following tables, constructed from details given by those 
authors, show to what extent the plants named were benefited 
by crossing either way. 








Primula veris (Wt. of seeds of 100 capsules 

) t62 

is to 



P. elatior (Av. No. of seeds per 

capsule) 46*5 




P. vulgaris 





„ var. alba [Scott] 





P. Sinensis „ 





„ [Hildebrand] 





P. Auricula [Scott] 





P. Sikkimensis „ 





P. cortusoides „ 





P. involucrata „ 





P. farinosa „ 





Hottonia pal. [Miiller] 





Pulmonaria o£E. [Hild.] 





Mitchella repens 





Linum grandiflorum 





L. perenne 





L. flavum (3 flowers produced capsules) 





* Forms, 

etc., p. 



The first observation is that in twelve cases the sliort- 
styled are in excess of the long-styled, and in four cases (f) this 
is reversed. Hence Mr. Darwin's conclusion is not absolute ; 
and it is a somewhat remarkable fact that Primula veris (the 
Cowslip) is the identical species from which he deduced the 
conclusion that the short-styled was the more feminine of 
the two forms. The conclusion now arrived at from this 
species would be, that when it is left to itself the short-styled 
form sets most seed ; but when artificially crossed it is the 
long-styled form which bears best. The cause of the former 
result is that some pollen in the short-styled form can fall 
upon the stigma and so secure self-fertilisation, which is 
impossible in the latter case. The same results occurred with 
Mr. Scott.* 

Hence Mr. Darwin's first conclusion, that the short-styled 
was the more feminine, was drawn from a wrong premise ; 
as it was not a question of sex so much as of union. When 
the results of self-fertilisation are compared, as given in the 
table on next page, it appears that the long-styled form of 
the Cowslip is the more feminine of the two, in the pro- 
portion of 42 to 30. 

Of that table, three cases of Frimula sp. (f) only show the 
short-styled bearing more seed than the long-styled when 
illegitimately fertilised ; viz., with Mr. Scott, P. vulgaris, var. 
alha, and P. Auricula (i.e. forms more or less modified by 
cultivation) ; and with Hildebrand, P. Sinensis, when crossed 

* Journ. Linn. Soc. Bot., vol. viii., 1864. This case may be taken to 
illustrate one of the disadvantages of £en accruing through great differenti- 
ation and adaptation to insect visitors. Though it appears proved that 
legitimate crossing sets most seed when carefully and artificially 
effected J yet, when the process is left to the capricious visits of insects, 
Mr. Darwin's experiments show how nature fails to derive the full 
benefit of intercrossing ; so that the Cowslip has to be contented with the 
results of the illegitimate union of the least fertile of the two forms. 


by distinct plants. The difference, however, being only two 
in each ease, is practically inappreciable. 

Of the other genera, Linum shows a slight inclination in 
favour of short-styled; but as this genus is exceedingly 
barren when illegitimately fertilised, the resnlts here given 
of that plant are insufficient for deducing conclusions ; at 
all events, these tables show that the long -styled form is 
certainly more prolific when illegitimately fertilised^ than the 
short-styled form when similarly treated.* 








Primula veris ( Wfc. of seeds of 100 capsules 

) 42 



P. elatior (Av. No. of seeds 






P. vulgaris 





„ var. alba [Scott] 





P. Sinensis „ 





„ [Hild.] (plants distinct) 





„ „ (same flower) 





P. Auricula [Scott] 





P. Sikkimensis „ 





P. cortusoides „ 





P. involucrata „ 





P. fariuosa „ 





Hottonia palustris [Miiller] 

(plants distinct) 





„ „ (same flower) 





Pulmonaria off. [Hild.] 


Mitchella repens 





Linum grandiflorum 





L. perenne 




' "ToQ^low"? 
Referring to the column of Differences in the first table, 
it will be noticed that two of the four marked (f) of the long- 
styled are considerable, namely, P. veris and Hottonia ; but the 

* Mr. Darwin noticed that this was the case with the genus Prinmla 
(Ix., p. 48). 


other two are practically inappreciable. On the other hand, 
considering every difference under 5 as inappreciable, there 
are four cases (§) of the short-styled in which it is consider- 
able ; and of these it was only 3 in the case of P. Sinensis with 
Hildebrand ; consequently one cannot confidently say that the 
short-styled is more feminine than the long-styled — at least, 
to any well-marked extent. 

With the corresponding column in the second table, one 
notices ni7ie cases where the difference is great ; while in all 
of those marked (f) it is inappreciable. Hence the con- 
clusion is much more pronounced in favour of the greater 
fertility of the long-styled forms when illegitimately crossed. 

Miiller accounts for "the greater productiveness of 
illegitimate crossings in the case of the long-styled form 
of Hottonia than in short-styled flowers, to the fact that the 
former kind of illegitimate crossings occur frequently in 
nature ; as these flowers are visited by pollen-seeking flies 
which have no need to thrust their heads into the flower 
of the short-styled form," which is, therefore, presumably 

The table I have here drawn up shows that the greater 
fertility of the long-styled form when illegitimately fertilised, 
is a general feature of heterostyled plants, and not peculiar 
to Hottonia pahistris ; hence we must look to a more general 

As another hypothesis, it may perhaps be suggested that, 
as the homomorphic condition of short stamens with a sliort 
style seems to have been the primitive form, then in the 

* If Muller's idea be true, Hottonia furnishes another instance of 
the disadvantage of great differentiations, and is only one degree better 
off than the Cowslip. In either case, one is inclined to ask what has 
become of the proper insects (whatever they may be) required for the 
perfect intercrossing of these flowers. 


long-styled form the stamens are uncTianged, while the pistil 
has elongated ; whereas, in the short-styled form, with now 
elevated stamens, these and their pollen have presumably 
become differentiated, while the pistil has remained un- 
changed. Now the above result appears to indicate the fact 
that the long-styled instil has not become physiologically 
differentiated to so great an extent as the pollen of the long- 
stamened form. The result is that it can be fertilised by the 
unchanged pollen of the same form more easily than the short- 
styled primitive form of pistil by the more highly differentiated 
pollen. This is not stated as a proved fact, and must be 
only regarded as a hypothetical suggestion. The extreme 
limits of differentiation are reached when the flower is 
heterostyled in form but dioecious in function. Thus 
j^gipMla ohdurata seemed to Mr. Darwin to be in a dioecious 
condition, but derived from heterostylism, in which the 
long-styled was apparently female, and the short-styled 

The species which shows the most marked difference 
between the produce of the legitimate fertilisation of the 
two forms is P. Auricula (or cultivated vars. of Auricula). 
It had been asserted by Prof. Treviranus that the long-styled 
unions were absolutely barren.* Mr. Scott shows that this 
idea arose from the fact that the plant in question had not 
been crossed. His experiments prove that the short-styled 
is the most fertile, whether legitimately or illegitimately 
crossed, though in the latter the difference is slighter : in 
the former the ratio being 8 to 6 ; and in the latter, 7 to G. 

Homostyled forms of P. Auricula are not uncommon. Mr. 
Scott found that 9 capsules gave 272 seeds, or an average 
of 30 seeds per capsule. Comparing this with the following 
results, its extreme fertility becomes apparent : — 
* Scott, I.e., p. 90. 


Shorfc-sfcyled X homostyled gave 8 seeds per capsule. 

Short-styled X short-styled „ 14 „ „ ,; 

Long-styled X homostyled „ 5 „ ,, „ 

Long-styled X long-styled „ 12 „ 
The pollen of the homostyled resembled that of the long- 
styled in appearance, though the stamens were situated high 
up as in the usual short-styled form. This seems to corrobo- 
rate what was said above ; for we have here also a long pistil 
fairly fertile with undifferentiated pollen. 

Another species of Frimula which often bears homo- 
morphic flowers is P. Sinensis. Mr. Darwin's attention was 
first directed to it by observing a long-styled plant — de- 
scended from a self-fertilised long-styled parent — with the 
stamens low down but with the pistil of the short-styled 
form, though the length of the style varied in different 
flowers on the same umbel. He fertilised eight flowers with 
their own pollen, obtaining five capsules with an average of 
forty-three seeds. The examination of the pollen of two 
equal-styled plants showed a vast number of small shrivelled 
grains. In the case of two whUe-Jioivered plants, in which the 
pistil was neither properly long-styled nor short-styled, the 
size of the grains was in the proportion of 100 to 88 ; whereas, 
between perfectly characterized long and short-styled plants 
it would have been 100 to 57. 

Of the first-mentioned homomorphic plants, four spon- 
taneously yielded 180 capsules, with an average of 54*8 
seeds, one containing 72 ; a result higher than could be 
expected of either form if self -fertilised. The next genera- 
tion proved to be all equal-styled, i.e. the grandchildren of 
the four original plants. One of these bore an average of Q'^ 
seeds per capsule, with a maximum of 82 and a minimum of 
40. Thirteen capsules, spontaneously self-fertilised, yielded 
an average of 532 seeds, " with the astonishing maximum, in 


one, of 97 seeds. In no legitimate union has so high an 
average of 68 seeds been observed bj me, or nearly so high 
a maximum as 82 and 97." * 

I give these results of homostyled. Auriculas and Chinese 
Primroses as illustrating the principle so abundantly proved 
amongst other plants — that as soon as they begin to retrace 
their steps from a prevailing differentiated condition self- 
fertilisation is rapidly resumed, and there follows a resumption 
of a vastly increased rate of seed-making. They prove, too, 
that however apparently stable these highly differentiated 
states may normally be, various conditions of environment 
can readilybreak them down ; thus, with cultivated plants, 
usually so much stimulated, starvation is a potent cause.t 
Linum perenne, as the above table shows, is particularly 
barren when illegitimately fertilised, but a single branch on 
a plant has been known to become homomorphic, and then 
to set seed abundantly; this occurred with Mr. Meehan. 
Warming found Menyanthes trifoliata to have become com- 
pletely homostyled in Greenland. 

Trimorphic Flowers. — As a type of heterostylism where 
a species adopts three forms, L. Salicaria may be taken. 
Briefly summarizing Mr. Darwin's elaborate experiments on 

* Forms of Flowers, pp. 218-221. 

t It is not only true with heterostyled plants, but the rule applies 
generally to highly cultivated flowers, that degeneracy from a floral point 
of view is correlated with enhanced powers of self -fertilisation. Thus 
a professional cultivator of Cyclamens is in the habit of keeping a stock 
of " worthless " weedy-looking plants, for the express purpose of raising 
seedlings, as they are so much more prolific than the true florists* types. 
Having obtained them, he then crosses them, and brings them up to the 
standard required. Indeed, the fact is well known to all cultivators, that 
the poorer the plant may be, from the florists' point of view, the better 
seed-bearer is it ; and that continually crossed and " perfect " flowers 
are proportionally impotent or tend to become so, when a tendency to 
become petaliferous often affects the essential organs. 


this plant, lie found that the flowers variously crossed gave 
the following results (omitting decimals under .6) : — * 


P.c. of 

Formed Number 


How crossed. 

capsules. of Seeds. 


... 38 


with mid-styled, 


. 51 


... 84 





. 107 


... 83 





. 81 


... 61 





. 65 


... 92 





. 127 


... 100 





. 108 


... t25 



long sta. of mid-st. 


. 55 


... 93 



long sta. of short-st. 


. 69 


... 54 



short sta. of 

long- St. 


. 47 


... to 



short sta. of mid-st. 


From these results Mr. Darwin conclnded that each form 
of pistil is as fully fertile as possible, only when it receives 
pollen from the stamens of the same length as itself, these 
being legitimate unions. It wdll be seen that the mid-styled 
form is the most fertile of the three when legitimately fer- 
tilised ; and as all illegitimate unions of the long- and short- 
styled forms were too sterile for any averages, the mid-styled 
form is also the most fertile when illegitimately crossed, and 
is least fertile with its own stamens, as indicated above by 
the (t). Hence self -fertilisation in this species is at a very 
low ebb. 

A few more remarks dednced from Mr. Darwin's observa- 
tions t niay be added here. From the three forms occurring 
in approximately equal numbers in a state of nature, and from 
the results of sowing seed naturally produced, there is reason 
to belief that each form, when legitimately fertilised, repro- 
duces all three forms in about equal numbers. 

When they are illegitimately crossed with pollen from 
the same form, they evince a strong but not exclusive tendency 
to reproduce the parent form alone. 

* Forms of Floxcers, p. 152. f L.c. p. 203. 


When the short or mid-styled forms were illegitimately 
crossed bj the long-styled, then the two parent forms alone 
were reproduced, but in no case did the third form appear. 

When, however, the mid-styled form was illegitimately 
fertilised by the longest stamens of the short-styled, the seed- 
lings consisted of all three forms. This illegitimate union was 
noticed as being singularly fertile, and the seedlings themselves 
exhibited no signs of sterility, but grew to the full height. 

Finally, of the three forms, the long-styled evinces some- 
what the strongest tendency to reappear amongst the off- 
spring, whether both, or one, or neither of the parents are 
long- styled. 

Although L. Salicaria has not, as far as I know, shown 
any signs of variability in the lengths of its filaments and 
styles, yet, as is perhaps generally the case with heterostyled 
plants, there are one or more species of the same genus which 
are normally homostyled. Thus L. hyssojpifoUum, which is 
not social, and is a dwarf form and an annual, bears only six 
to nine stamens, the anthers of which surround the stigma, 
which is included within the calyx. The three stamens, 
which vary in being present or absent, correspond with the 
six shorter stamens of L. Salicaria. The stigma and anthers 
are upturned as in the last species, and so indicate the fact 
that it is a degenerate form from L. Salicaria or some other 
intercrossing species, though it has now reacquired its self- 
fertilising properties. Oxalis is a genus having trimorphic 
species. Many of them are extremely infertile with their 
" own form " pollen. Such are the long-styled form of 0. 
tetrapJiylla, versicolor, Brasiliensis, and compressa. On the 
other hand, in the long- styled form of 0. incarnata, rosea, 
and Piottce, and in the mid-styled form of 0. carnosa, no self- 
sterility occurs.* 

* According to Hildebrand, Bot. Zeitg., xlv., pp. 1, 17, 33. 


Origin of Heterostylism. — The question may be now 
asked, How has heterostylism arisen ? We have seen, in the 
first place, that in many cases there is a certain instability in 
the length of the filaments of the stamens and of the styles, 
in that they are liable to alter spontaneously, and especially 
under cultivation.* In the case of Fritnula Auricula, the 
homomorphic form has the anthers and stigma at the orifice, 
while in P. Sinensis they are often both low down ; it is clear 
that either might arise in two ways. In the case of the former, 
the stamens, while resembling in position that of the stamens 
in the short-styled form, have pollen like that of the long- 
styled, the pistil being of that kind. Hence it is reasonable 
to assume that the anthers have been uplifted. In the 
Chinese Primrose it is the reverse ; so that the pistil 
of a long-styled form has been loivered to the level of the 
stamens; the stigmas, too, are that of the short-styled. 

Recognizing this instability of the essential organs, it is 
reasonable to assume that it may be due to varying degrees 
of nutrition which can readily bring about such changes, a 
relatively strong vegetative vigour elevating the stamens in 
the one case, while a slight tendency to degeneracy with 
lessened vital vigour tends to suppress the pistil in the other. 

Assuming a homomorphic form to have been the primitive 
and ancestral state, we can realize how dimorphism has been 
brought about by such varying degrees of stimulus having 
been applied to the stamens and pistil. Insect agency 
I take to have been this cause, which, at the same time, has 
by selection fixed the heights of the stamens and style so 

* See the description, given above, of Narcissus co^uus, Fig. 37, 
p. 121. Mr. Darwin found Gilia to vary much in this respect. It may- 
be added that it is a not uncommon featiu'e in flowers which are not 
heterostyled, as e.cj. cultivated Gladioli and CVoci, Fritillaria Meleagris, 


as to reader them permanently dimorphic for legitimate 
fertilisation. The predominant insect or insects were (as I 
surmise) the direct cause of arresting the fluctuations which 
they themselves, as well as accidental sources of nutriment, 
had set up in the lengths of the essential organs, thus 
compelling them to retain their anthers and stigmas at the 
correct height. 

If there were from one to three prominent kinds of 
insect-visitors the flowers might become adapted to them, 
and trimorphism be the result ; if four, tetramorphism ; and 
there is no a priori reason why there should not be polymor- 
phic flowers as well, in the strict sense of the prefix of that 
tei'm, provided a flower could furnish a sufficiency of stamens. 

It is further to be noticed that the rule holds good with 
heterostyled plants, as with all other kinds of differentiation, 
that in nature, whenever self-fertilisation can be effected, 
more seed is borne than by the forms requiring intercrossing. 
First, whenever it can be brought about mechanically ; as 
has been observed in P. Sinensis, by the corolla, when falling 
off, dragging the anthers over the stigma in the long-styled 
form, which consequently yields more seed.* In P. veris, it 
does not do so ; but as pollen can fall in the short-styled 
form, in this species that form is thus the most fertile (see 
above, p. 205). 

Secondly, when these plants are artificially and legiti- 
mately fertilised, and not left to the chance visits of 
capricious insects, then the results are as they should be ; 
but if self-fertilisation be artificially and repeatedly practised, 
then nature responds to the act ; the anthers and pollen may 
in part degenerate, but what is left good is ample to secure 
abundant seed, and the self-fertilised form surpasses even the 

* Darwin found that, in the absence of insects, the long-styled form 
of P. Sinensis was twenty-four times as productive as the short-styled. 


legitimately fertilised heteromorphic unions in fertility. 
Thus, Mr. Darwin observes, "The self -fertility of Primula 
veris increased after several generations of illegitimate fertili- 
sation, which is a process closely analogous to self -fertilisa- 
tion." * 

Lastly, if homomorphic forms occur spontaneously, as is 
often the case with species of Primula, Mr. Darwin has 
shown they are not only " capable of spontaneous legitimate 
fertilisation, but are rather more productive than ordinary 
flowers legitimately fertilised." f 

It was Mr. Scott who suggested that the equal-styled 
varieties arose through reversion to a former homostyled 
condition of the genus. Mr. Darwin supported this view in 
consequence of observing " the remarkable fidelity with which 
the equal-styled variation is transmitted after it has once 
appeared." J 

* Cross and Self Fertilisation, p. 351. 

t Forms of Floiuers, p. 273; and Cross and Self Fertilisation, p. 352. 

X Forms, etc., p. 274; Mr. Darwin was so profoundly impressed wdth 
the supposed advantages of intercrossing, that he again and again 
asserts most positively that self-fertilisation is injurious, often in 
diametrical opposition to his own statements and experiments. Thus, 
while speaking of heterostyled trimorphic plants, he says, " As I have 
elsewhere shown (The Effects of Cross, etc.), most plants, when fertilise(5 
•with their own pollen, or that from the same plant, are in some degree 
sterile, and the seedlings raised from such unions are likewise in some 
degree sterile, dwarfed, and feeble." Yet, in the work quoted, he has 
not only shown that, when he persevered with self-fertilisation for several 
generations, he found it was just the reverse; as e.g. with "Hero" Ipomceat 
the white Mimulus, etc., and with Primula, as stated above ; but he 
more than once draws an opposite conclusion, as when speaking of self- 
fertile varieties (I.e., p. 352): *'It is difficult to avoid the suspicion 
that self-fertilisation is in some respects advantageous. . . . Should this 
suspicion be hereafter verified, it would throw light on the existence [of 
cleistogamy]." It is this " suspicion " which I have completely veri- 
fied ; and, indeed, any idea of " injuriousness " is refuted by the majority 


Besides the more obvious differences in the relative 
lengths of the styles and filaments* of heterostyled flowers, 
the rule is for the stigmas of the long-styled to be larger or 
longer than those of the short-sfcyled,t and to have their 
papillae longer and broader. 

Thus in the nine species of Primula described by Mr 
Darwin, in two only were the stigmas nearly alike in both. 
Of three species of Linum, X. jiavmn alone had an appre- 
ciable difference in the stigmas. In Pulmonaria officinalis 
and Polygonum fagopyrum, Forsythia suspensa and JEgipMla 
elataj it was not, or scarcely appreciable. 

Again, besides those mentioned there were twenty species 
in which the stigmas of the long-styled were markedly 
superior to those of the short-styled. 

of plants in a wild state being constantly self-fertilised, as Miiller, 
and, indeed, Mr. Darwin himself has shown to be the case. Thus, 
he gives two lists, of forty-nine species in each, (Cross and Self Fert., 
etc., pp. 357 and 365), one of self-sterile, the other of self-fertile plants, and 
adds, " I do not, however, believe that if all known plants were tried in 
the same manner, half would be found to be sterile within the specified 
limits; for many flowers were selected for experiment which presented 
some remarkable structure ; and such flowers often require insect aid " 
(I.e., p. 270). The proportion of self-sterile plants is, in fact, extremely 
small. Miiller remarks, e.g., of the highly differentiated order Scrophu- 
larineoe^ that " in default of insect-visitors, self -fertilisation takes place 
in most forms ; and in only a few are insect- visits, and consequently 
cross-fertilisation, so far insured that self-fertilisation is never required 
and hag become impossible." Similarly of LabiatcB he says, " Self- 
fertilisation seems to be rendered impossible only in the species of 
Nepeta, Thymus, Mentha, and Salvia described " (Fertilisation, etc., pp. 
4G4 and 503). Moreover, while Mr. Darwin includes the Fox-glove and 
Linaria vulgaris among his sterile plants, Miiller considers them both 
to be self -fertilising. 

* Exceptions occur, thus Cordia and Linum grandijlorum have little 
or no difference in the length of the stamens. 

t Leucosmia Burnettiana is remarkable for having the stigma of the 
short-styled form the more papillose (Forms of Flowers, p. 114). 


Oa the other hand, the anthers of the short-styled are 
usually longer and contain larger pollen grains than those 
of the long-styled, the pollen of which is also often more 
translucent and smoother. 

Of all the species included in the above-mentioned thirty- 
six species, only five seem to have the pollen of both forms 
of the same size, and two in which it was reversed. The 
five species are Leucosmia Burneitiana, Linum grandifloruvn, 
Cordia, Gilia pulchella, and Coccocypselum. The two in which 
the pollen grains of the long-styled form were the larger, 
were Gilia micrantha and Phlox subulata. 

The presence of cases where the usual differences are not 
pronounced is just what one expects to find, in accordance 
with the laws of differentiation ; w^hereby intermediate 
conditions are to be looked for. Thus some species of 
Primula afford great differences in the shapes of the stigmas, 
P. veris being globular in the long-styled, and depressed 
in the short-styled; while in P. Sinensis it is elongated: 
but in other species, as P. Sikkimensis and P. farinosa, there 
is but little difference between the stigmas of the two forms. 
In some cases the differences reside entirely in the stamens or 
pollen grains, as in Forsythia suspensa, in which, although 
(contrary to the rule) the anthers of the long-styled are in 
length as 100:87 compared with the short-styled, yet the 
pollen grains are as 94* : 100, which agrees with the rule. 
With. Liiium grandifiorum and Cordia and Gilia pulchella, etc., 
the difference lies in the pistil. On the other hand, the 
difference may reside in the stamens, as in ^giphila elata, 
the pollen grains being as 62 : 100, i.e. in the long-styled as 
compared with the short-styled. 

j^giphila ohdurata has the stigmas of the long-styled 
in length 100 : 55 as compared with the short-styled ; and 
the length of the anthers as 44< : 100. This is, therefore, 


apparently trnlj heterostyled, but from Mr. Darwin's obser- 
vations be tbinks tbe sbort-stvled incapable of fertilisation ; 
moreover the anthers of tbe long-styled form were " browm, 
tougb, and devoid of pollen." He considers that, from baving 
been beterostyled, it has now become dioecious, or else gyno- 

M. W. Durek has sbown * tbat several genera of 
Buhiacece are beterostyled in form but quite dioecious. 

Faramea affords anotber curious difference. In tbe long- 
styled form tbe stigma is sbort and broad ; in tbe sbort- 
styled, it is long, thin, and curled. Tbe anthers of tbe 
sbort-styled are a little larger than tbose of tbe long-styled, 
and tbe size of their pollen grains are as 100 : 67. But tbe most 
remarkable difference (of wbicb no otber instance is known) 
is in tbe fact tbat wbile tbe pollen grains of tbe sbort-styled 
forms are covered witb sbarp points, tbe smaller ones are 
quite smooth. Tbe anthers, moreover, rotate outwards in 
the short-styled, but do not do so in tbe long-styled flowers. 
A similar rotation takes place in some of tbe Cruciferce, and 
facilitates intercrossing. A somewbat analogous torsion 
occurs in some styles and stigmas, as of Linum jperenne, 
Luzula arvensis, Begonia, etc. 

Tbe smaller and smootb pollen, in tbe more degenerate 
condition of tbe long-styled form, is suggestive of the origin 
of tbat of wind-fertilised flowers, wbicb bas sometimes 
acquired tbe same form. Indeed, tbe two forms of pollen 
(figured by Mr. Darwin at p. 129 of Forms of Flowers) exactly 
correspond to tbe very common spinescent form in inter- 
crossing species of Comijositoey and to tbat of tbe anemopbilous 
Artemisia of the same order, respectively. 

Tbe general conclusion, therefore, derived from tbe com- 

* Sur V Organisation Florale chez quelques Ruhiacees. Ann. Jard. 
Bot. Biiitenzorg 3, p. 105. 


parison of these minute details, is that the long-stjled form 
of flower represents a more fully developed pistil, and 
therefore a more female condition ; while the short-stjled is 
more male : and, as we have seen above, this is borne out by 
the comparison of the offspring ; and, lastly, by the probable 
dioecious condition of ^gijphila ohdurata, as well as by the 
actual dioecism of some species of Musscenda and Morinda 
umhellata ; while Musscenda cylindrocarjpa and certain other 
species of Morinda are hermaphrodite without heterostylism 
(Burck, I.e.). 




Gynodicecism and Gynomon(i:cism. * — In accounting for the 
origin of certain floral structures, it must be borne in mind 
that the habits and constitutions of plants are so infinitely 
various, that the interpretation given for that of a structure 
in one case may fail to be satisfactory when tested by 
another ; and an argument apparently sound for the expla- 
nation of a special phenomenon in a particular plant «r 
plants may not at all apply to that of others. Thus, while 
the Hazel may mature its stamens before the pistils on a 
slight rise of temperature in early spring, there are many 
herbs, if they happen to blossom in spring earlier than is 
their custom, in summer, or what may be their optimum 
period, may have the staminal whorl more or less deranged, 
as such plants require a relatively higher temperature to 
develop them perfectly.f This is particularly characteristic 
of gynodioecious plants. Thus, e.g., most of the distinctly 
protandrous species of the Alslnece are in this condition, and 

* Gynodiajcinn signifies that the same species may have both female 
and hermaphrodite plants. 

Gynomoncecism signifies that the same plant may bear both female 
and hermaphrodite flowers. 

t This will be discussed more fully in the next chapter. 


the plants with small, usually pistillate flowers are chiefly 
in blossom at the beginning of the flowering period of tbe 
larger-flowered hermaphrodite plants of this section of the 
Caryophyllece. Similarly, Caffea arahica produces small pis- 
tillate flowers in Guatemala at the beginning of the season. * 
It is the same with Geranium macrorhizon and many species 
of Pelai-gonium, etc. f 

Gynodioecism also prevails in the Lahiafce, but both female 
and hermaphrodite plants for the most part blossom simul- 
taneously in summer. It may be noticed that the corolla is 
almost invariably reduced in size in female flowers, whether 
the species be strictly dioecious as in Bryony, or gjmodioecious 
as Thyme, showing the close interdependence between the 
corolla and stamens. X 

That climatal conditions are likewise connected with 
the Gynodioecism of the Lahiatce seems probable from the 
behaviour of Thymus Serpyllum ; for Delpino found that it 
was trimorphic in the warmer region of Florence, having 
flowers with greatly developed stamens and the pistil in 
every stage of abortion or even absent (see Chapter XXV.) ; 
other flowers showed the exact converse ; and, lastly, others 
were hermaphrodite. IMiiller, however, on the other hand, in 
Westphalia and Thuringia; Ascherson, in Brandenburg; 
Hildebrand, in the Rhine provinces; and Mr. Darwin, in 
England, never met with the purely male form ; though 
Dr. Ogle found some with the pistil permanently immature. § 
Similarly, Eriophorum angustifolium is gynodioecious in 
Scotland and the Arctic regions. || 

Besides temperature, the character of the soil has most 
probably much effect in bringing about this kind of partial 

* Miiller, Fertilisation, etc., p. 304. f L.c, p. 158. 

X See Forms of Flowers, pp. 307-309. 

§ Miiller, I.e., p. 474. U Forms of Flowers, p. 307. 


diclinism. Mr. Darwin thought " a very dry station 
apparently favours the presence of the female form,"* i.e. 
a lessened vegetative vigour tends to check the development 
of th^ corolla and stamens, especially if a low temperature 
accompanies it ; just as, conversely, we have seen how a 
high temperature enhances it. Mr. Hart thus found that, 
with Nepeta GlecJioma, all the plants which he examined near 
Kilkenny were females; while all near Bath were hermaph- 
rodites, and near Hertford both forms were present, but 
with a preponderance of hermaphrodites. f 

Both Miiller and Mr. Darwin offer theories to account for 
the origin of these gynodicecious plants. 

Miiller, after quoting Hildebrand's view, which he rejects, ;J; 
says,§ " Of the flowers of the same species growing together, 
the most conspicuous are first visited by insects, and if the 
flowers on some plants are smaller than on others, perhaps 
owing to scanty nourishment, they will generally be visited 
last. If the plant is so much visited b}^ insects that cross- 
fertilisation is fully insured by means of protandrous dicho- 
gamy, and self-fertilisation is thus rendered quite needless, 
then the stamens of the last-visited small-flowered plants 
are useless, and Natural Selection wdll tend to make them 
disappear, because the loss of useless organs is manifestly 
advantageous for every organism. 

" This explanation rests upon the hypotheses, (1) that 
the flowers of those sjDCcies in which small-flowered female 
plants occur together with large-flowered hermaphrodite 
plants are plentifully visited by insects and are markedly 

* Forms of Floivers, y). 301 

t NaiuYC, 1873, p. 1G2 ; and see below, p. 239. 
X Fertilisation, etc., p. 473. 

§ L.c, p. 484. Compare his- remarks on Scahiosa arvensis, I.e., pp. 
310, 311. 


protaudroTis ; (2) that variation in size of the flo^ve^s has 
always taken place, not among the flowers on a single plant, 
but between the flowers on different individuals." 

Mr. Darwin suggests another view : * "As the production 
of a large supply of seeds evidently is of higb importance 
to many plants, and . . . the females produce many more 
seeds than the hermaphrodites, increased fertility seems to 
me the more probable cause of the formation and separation 
of the two forms.'* 

" S. M.," reviewing Mr. Darwin's work in the Journal of 
Botany, 1877, p. 375, "felt compelled to differ from the 
author, and adds, " For ourselves we cannot help thinking 
that gynodioecism can be better explained on the view of a 
sufficiency of pollen for the fertilisation of all the individuals 
of a species being produced by only a few of the flowers, 
so that instead of some of the anthers of all the flowers 
becoming abortive — a very common occurrence — we see here 
abortion of all the anthers of some of the flowers. ... All 
known instances of gynodioecism relate to species which 
have the maximum of stamens possessed by the orders to 
which they relatively belong, and are without any complex 
entoraophilous structure. . . . We may also remark on the 
pauciovulate condition of gynodioecious species, and ask why 
do we not see this form of sexual separation in multiovulate 
ones ? " 

In reply to this writer's suggestions, I would remark 
that in all entomophilous flowers far too much pollen is 
produced and wasted ; that Mr. Darwin's observation, that 
a bee could fertilise ten pistils with pollen from one flower 
of Satiweia, might readily apply to hundreds of cases where 
no gynodioecism exists ; and as long as insects visit flowers 
the tendency is not to contabescence and abortion of the 
* Forms of Flowers, p. 30-i. 


anthers, but to higher differentiations and an increase in the 
quantity of pollen. Secondly, that the orders, with gyno- 
dioecism have the maximum of stamens, is not universally 
true, Pelargonmm having only seven out of ten. Again, the 
Lahiatce are especially characterized by " entomophilous 
structures." Lastly, the order Caryophyllece is multiovulate. 
In the first two interpretations, those of Miiller and 
Darwin, Miiller suggests scanty nourishment as a cause for 
the diminished size of the female flowers, which might apply 
to any or every protandrous plant and so give rise to gyno- 
dioecism ; for if it be a sufficient cause in one family, why has 
it not brought it about in all ? This cause alone does not 
touch the question, Why is gynodioecism peculiarly common 
in the Alsinece of the CaryophylletE and in Lahiatce ? Mr. 
Darwin thinks that an increased fertility of the female may 
be the cause ; but he seems to forget that no flower of the 
Lahiatce can bear more than four seeds, so that, supposing a 
female plant to have the same number of flowers as a her- 
maphrodite, if it bears more seeds it must be due to the 
decrease in fertility of the latter, and not to any increase in the 
former. * It is, in fact, a very common occuiTence for a floAver 
of any member of the Lahiatce to bear one, two, or three only, 
as Avell as four nutlets in an individual fruit. Mr. Darwin 
" doubts much whether natural selection has come into play," 
and notices that " the abortion of the stamens ought in the 
females to have added, through the law of compensation, to 
the size of the corolla," as is the case in the ray florets of 
the gynomoncccious Composites. He, however, recognizes the 

* In his experiment with Satureia hortensis, Mr. Darwin collected 
seeds from the finest of ten female plants, and they weighed 78 grains ; 
while those from the single hermaphrodite, which was a rather larger 
plant than the female, weighed only 33"2 grains ; that is, in the ratio of 
100 to 43 {Forms of Flowers, /p. 303). 


intimate connection between tlie corolla and androecium, and 
thinks that "the decreased size of the female corollas is due 
to a tendency to abortion spreading from the stamens to the 

In noting all the plants mentioned by Miiller and Mr. 
Darwin as gynodioecious, there are besides the two well- 
marked groups already mentioned, viz., AlsinecB and Labiatce, 
the following isolated genera or species, Pelargonium, Gera- 
nium macrorliizon, Sherardia arvensis, Valeriana montana, 
Scahiosa, Cnicus, Ecliium vulgar e and Plant ago ; to the 
Compositoi, I can add Achillcea millefolium; and I think 
also Vines may be included in the list. 

The first and important point to note about the flowering 
of the Alsinece is that the female flowers are the first to open, 
at the hegimiing of the season* It is the same with Geranium 
macrohizon, Pelarguniuni, and Coffee in Guatemala. Now, 
we have already seen how sensitive the androecium and the 
corolla are to a low temperature, so that we have here a direct 
cause which will account for the check upon the growth and 
development of these two whorls. Applying this principle 
to the Labiatce, we must remember that as a group they are 
correlated to a warmer climate, their " home " being the 
Mediterranean and even warmer regions ; hence I assume 
their greater hereditary sensitiveness to a low temperature 
in those descendants which occupy a cooler temperate zone. 
This may, I think, account for the predominance of purely 
female forms, as well as the presence of stamens in every 
degree of degeneracy. 

How far the same principle will apply to the other 
gynodioecious genera and species, I will not pretend to 
oiler an opinion, as not enough is yet known about them ; 

* See Hildebrand's observation, p. 234, and Sexuality and Tempera- 
ture, p. 237. 


only we must always remember that there may be a variety 
of causes which may equally well bring about the same 

It may be also borne in mind here that another result 
of low temperature is, while retaining the function of the 
androeciura, to arrest the expansion of the corolla and to 
render the flowers self-fertilising. This is peculiarly the 
case with the Alsinece ; while Lamium amplexicaule fails to 
open its earliest small- flowered flowers at all, being strictly 

The preceding cases of gynodioecism are all associated 
with a more or less degree of protandry. It is rarer to find 
it accompanied with protogyny in the hermaphrodite form. 
Miiller records it in Plantago lanceolata in England, which I 
can corroborate, and in P. media in Germany. These plants 
are anemophilous, and in a state of passage from an ento- 
mophilous ancestry ; s -^ that it may have been retained from 
an early condition. 

Gynomonoecism is not particularly common, except in the 
Compositm, where the ray florets are often female, while 
the disk florets are hermaphrodite. This is due to com- 
pensation ; for transitional states may be seen in flowers 
which are passing into the "double" condition; for as the 
corolla changes its form and becomes ligulate, the stamens 
are suppressed, and the style arms alter their shape. 
Anemone hepatica is said to be gynomonoecious,* and also 
Syringa Persica.f 1 have seen no case, and no description is 
given of these two, so that I can only suggest that it may 
be a result from degeneracy, perhaps on the road to a 
petaloid condition of the stamens. Such a state I have 
found in a Plantago which was gynodicecious. 

* Dr. S. Calloni, Arch. Sci. Phijs. et Nat, xiii., 1885, p. 409. 
t Miiller, Fertilisation, etc., p. 393. 


Androdicecism and Andromoncecism.* — These conditions 
do not appear to prevail to the same extent as the female 
forms of flowers. Both of these kinds are not at all un- 
common in the UmhelUferce, and are a result of exhaustion, 
for the umbels produced at the end of the season are often 
entirely male ; or, if at other periods, it is generally the 
central florets which develop no pistils, as in Astrantia minor. 
Miiller has noticed how " the weaker plants usually bear but 
one umbel consisting only of male flowers." This would 
make it androdioecioas. I find that andromonoecism prevails 
in Astrantia major ^ Carum, Smyrnium, and in Trinia vulgaris. 
This last, growing on the Clifton downs, bore umbels which 
were altogether male, after the hermaphrodite ones had 
formed their fruit. Daucus grandiftora is remarkable for 
having three kinds of flowers. According to Miiller, the 
central ones are male; at the edge of the umbellule the 
flowers are neuter, with the outermost petal greatly enlarged; 
lastly, at the margin of the whole umbel, are female florets 
in which the outer petals attain to a gigantic size.f 

* Androdicecism signifies that the same species has both male and 
hermaphrodite plants. 

Andromonoecism signifies that the same plant bears both male and 
hermaphrodite flowers. 

t I would here remind the reader that the interpretation given 
above (Chapters XI.-XIII.) of the origin of irregular corollas, applies 
equally well to those cases where it is only m the outermost florets of a 
cluster where the petals are enlarged, as in Iheris, many of the Com- 
'positce, and UmhellifercE, as well as in Hydrangea, Guelder Rose, etc. In 
all these, when insects first approach the umbel and alight on the border 
of it, any or each individual floret on the margin may have to carry the 
burden. As soon, however, as the insect passes the edge of the cluster, 
its weight is distributed over several florets ; so that they are not sub- 
mitted to any special strains upon one, i.e. the outer side only. The 
same remarks apply to Mentha, as compared with Lamium. The insect 
visits one flower at a time in the latter, but scrambles over several in 
the former, which has (presumably) degraded in consequence. 


Caltlia palustris is said to be androdioecious, but no 
details are given by the observer.* 

Besides the JJmhellifera!,\ where andromonoecism seems 
to be a characteristic feature, Miiller mentions Asperula 
taurina and Galium Cruciata, Pidmonaria officinalis^ Coriaria 
myrtifolia,ai\d Biospyrus Virginiana as being andromonoecious. 
The hermaphrodite flowers of these species are protandrous. 

In Galium Cruciata^ Mr. Darwin noticed that the pistil is 
suppressed in most of the lower flowers, the upper remaining 

Heterostylism may tend to produce the same result when 
the stamens of the long-styled forms degenerate so far as to 
become atrophied without the pistil losing its functions. 
Pidmonaria angustifolia and Phlox suhulata give hints of this 
condition,;!: Asperida scoparia was at first thought by Mr. 
Darwin to be heterostyled, but finding the anthers to be des- 
titute of pollen, he considered it to be dioecious. A. taurina, as 
figured by Mii]ler,§ shows great variability in the lengths of 
the filaments and styles, and he pronounces it to be andro- 
monoecious. Hence, as so many of the Euhiacece are hetero- 
styled, there seems every probability of one result of this 
peculiarity, being one or other kind of this incompletely 
affected or partial diclinism. In the case of Coriaria myrti- 
folia, Hildebrand found that it was the first flowers which 
were male only. In Maples, as in Galium Cruciata, the rule is 
for the three or more flowered corymb to have the central 
one hermaphrodite, and the lower or outer ones male. This 

* Lecoq, Geog. Bot., torn, iv., p. 488. 

t Miiller says that in Sanicula Europcea the outer flowers are male, 
and develop after the inner ones, which are hermaphrodite. This is so 
anomalous, that one suspects an error somewhere. I have not hud any 
opportunity of examining fresh flowers. 

X Forms of Flowers, p. 287. 

§ Fertilisation, etc., p. 303. 


clearly is a question of the distribution of nutrition ; the 
lou-er, being the later ones to expand, are the weaker.* 
Miiller mentions Horse-chestnuts as being also andro- 
monoecious ; and what is exceptional is that the hermaphro- 
dite flowers are protogjnous. This, however, may be due 
to the early period of flowering, like species of Prunus and 

The reader will now perceive that there may be several 
causes at w^ork to produce these kinds of " partial diclinism;" 
and that what is required is to ascertain, if possible, by 
observation and experiment, which is the one peculiar to 
each species. Secondly, when any one or more causes has 
been sufficiently persistent, the results become hereditary ; 
so that certain species, genera, and orders become more or 
less characterized by these peculiar features. 

* Compare the observations on Adoxa, p. 188. 





General Observations. — As the environment is now known 
to have most potent influences on the anatomical structure 
of the vegetative system of plants, thereby affecting their 
outward and visible morphological characters as well ; so are 
there many causes which affect the reproductive system, at 
one time influencing the androecium, at another the g^noe- 
cium, favouring them or the reverse as the case may be ; so 
that eitlier sex or even both may be entirely suppressed, and 
a hermaphrodite flower become male, female, or neuter. 

With regard to the most general agency, there seems to 
be a tolerably uniform consensus of opinion that the female 
sex in plants is correlated with a relatively stronger vital 
vigour than the male ; and this is just what an a 'priori 
assumption would look for, as the duration of existence and 
the work to be done in making fruit require a greater 
expenditure of energy than the temporary function of the 

We must, however, distinguish between a healthy vital 
vigour, and any excessive vegetative growth, as occurs under 
high cultivation, and as is often the result of intercrossing. 
If this latter surpass the requisite or optimum conditions for 
the healthy performance of the functions of all the organs 


of a plant, then eitlier of the sexual organs mav begin 
to deterioriate, till they become metamorphosed into petals 
or leaves, or else degenerate and vanish. 

It is true enough that we know nothing of the real nature 
of life ; but it is easy to see that, of the various phases of 
development, from germination to the production of seed, 
each should have the proper amount of energy at its disposal, 
and no more ; for if any one organ be stimulated beyond the 
optimum degree, others suffer through atrophy. The first 
and well-known distinction to be noticed lies, of course, 
between the " vegetative energy," by means of which 
roots, stems, branches, and foliage are developed, and the 
"reproductive energy," which brings about the formation of 
flowers, fruit, and seed. If either of these be unduly excited, 
the other diminishes. Thus, as long as fruit trees are 
developing much Avood and foliage, they either bear fruit 
badly or not at all. Plants which are propagated largely 
by vegetative means of multiplication, such as bulbs, corms, 
tubers, etc., are notorious for failing to set seed as well. As 
an instance in nature, Ranunculus Ficaria may be mentioned. 
This plant propagates itself by " root-tubers " and by aerial 
corms, and rarely produces much fruit, for the pollen often 
remains in an arrested state.* Conversely, if vegetative 
energy be checked by root and branch pruning, bark-ringing, 
etc., the reproductive energy is promoted, and an abundance 
of fruit is the reward. Similar results follow a decrease of 
energy through impoverishment, when enormous crops of 
fruit may be borne by trees, as I have seen in Portugal 
Laurels, when the roots had penetrated a bed of gravel and 
the branches became decayed. 

Apart from these general considerations certain special 
conditions are found to favour one sex more than the other, 
* See Van Tieghem on R. Ficaria, Ann. des Sci. Nat., v., ser. 5, p. 88. 


so that normally hermaphrodite flowers may become uni- 
sexual, and every possible degree between these two extreme 
cases can be met with in nature and cultivation. The 
problem, therefore, is to what the immediate causes 
may be in each case which stimulate or suppress the energy 
required for the proper development of the stamens and 
pistil respectively. 

There appears to be a closer bond between the stamens 
and corolla than between the two kinds of essential organs 
themselves ; * thus, if the corolla degenerate, the antipetalous 
stamens at least tend to follow, as in the Alsiiiece. On the 
other hand, the first tendency towards " doubling " appears 
in a more or less pronounced petalody of the andrcecium. 

As petals are a nearer approximation to foliar organs, 
the above means that vegetative energy is more prone to 
affect the stamens, when from some cause they have first 
becun to lose their proper function, than the pistil. 

The pistil may fail in its development from two classes 
.of causes: either from an undue display of the vegetative 
vio"our, as in completely double flowers — though it may be 
unaffected in a partially double one ; or else from excessive 
feebleness, under which a flower may succeed in making the 
andrcecium, but has not sufficient energy to develop the 
gynoecium ; as, e.g., often takes place in the flowers of the 
Umhelliferce at the close of the season. 

There is no absolute rule in these matters, and differences 
result from various degrees of energy at the disposal of the 

♦ A study of the vascular system of flowers and their axes bears this 
out, as the provision made for the stamens usually arises from the 
perianthial cords, while that for the pistil is mostly isolated off in rather 
a more marked and independent manner. Exceptions occur, as in 
Ballota nigra, in which the four stamens originate from the same corda 
as those of the placentas. 


whorls, giving rise to corresponding results of different 
degrees of development in the respective sexes. 

The points to be clearly perceived are that a plant should 
be able to develop all its organs in perfection; that there is 
an optimum degree of energy for each ; and that, though it 
is customary to group these energies under the two expres- 
sions, vegetative and reproductive, yet the principle may be 
carried out in detail : so that, e.g., an enlarged corolla tends 
to destroy the stamens, as of the ray florets of Dahlia, or even 
the pistil too, if it be very large, as in Centaurea. A stimu- 
lated androecium brings about an arrest in the pistil, and 
causes protandry ; and if the perianth be highly developed, 
as in orchids, the enhancement of the former may cause 
degeneracy in the ovules. 

Sexuality axd Nutrition. — Assuming, for the present, that 
the ancestral condition of all flowers, excepting, perhaps, those 
of the Gymnosperms, was hermaphrodite, many instances 
exist of the same species having male, female, and herma- 
phrodite flowers, such as the Ash, Silene inflata, etc., where 
the aborted organs often remain more or less rudimentary. 
It cannot be pretended as yet that the cause or causes can 
be at all positively asserted, in each case, for the tendency 
to abortion either in the stamens or pistil ; but there are 
certain well-ascertained facts which can undoubtedly play a 
part in the processes of degeneration or exaltation of the 
stamina! and carpellary energies respectively. If they be 
sufficiently persistent the subsequent generations can, then, 
become completely diclinous, without a trace of the other 
sex remaining ; yet, as is well-known, any diclinous plant 
may reproduce by reversion the lost sex, thereby revealing 
its original hermaphroditism. 

In endeavouring to trace the present condition of 
diclinous flowers back to an ancestral hermaphrodite condi- 


tion, it will be as well to consider certain signiBcant facts 
which may help ns in ascertaining the cause of their present 

" Hildebrand has shown," writes Mr. Darwin, " that with 
hermaphrodite plants which are strongly protandrous the 
stamens in the flowers which open first sometimes abort, . . . 
Conversely the pistils in the flowers which open last sometimes 
abort." Similarly Gartner observed that "if the anthers on a 
plant are oontabescent (and when this occurs it is al\va3^s 
at a very early period of growth) the female organs are 
sometimes precociously developed." * 

A reason for this is that, on the one hand, since a higher 
temperatures is correlated with protandry, the first flowers open 
when the o])timum temperature has not arisen; so that the 
stamens are checked, a cooler temperature being less inimical 
to the development of the gynoecium. On the other hand, 
the last flowers of the seasop. are produced when the vital 
energy is waning, and although the flowers may expand, they 
are too feeble to develop the pistil. 

Now exactly the converse may occur; thus Mr. W. Gr. 
Smith called attention f to the seemingly unobserved fact 
that Euphorbia mnygdaloides always bears terminal male 
flowers alone at first, and subsequently the two sexes together 
on lower lateral " flowers." This agrees with Castanea 
Americana,^ as noticed by Mr. Meehan. In these two cases, 

* Forms of Flowers, p. 283. I hardly think this can be always the 
case ; for, of Vines growing side by side, some will occasionally have 
the anthers utterly devoid of sound pollen, but with the pistil normal; 
while others will be entirely hermaphrodite with no sign of contabescence. 
I have examined such, supplied to me by Mr. Barron from the gardens 
of the Royal Horticultural Society at Chiswick. The cause is at present 
veiy obscure. 

t Journ. ofBof.y 1864, p. 196. 

X Proc. Acad. N. Sci. of Philadel., 1873, p. 290. 


therefore, we have instances of the plants flowering and 
bearing male organs only before the highest effort of vital 
energy is displayed — the preliminary and feebler effort being 
capable of developing the androeciiim alone. 

With regard to diclinous trees, many examples could be 
found to illustrate tbe principle that the female flowers are 
normally produced by stronger shoots than the male. Mr. 
Meehan has particularly called attention to this fact. For 
instance, " Jicglans nigra* exhibits three grades of growing 
buds. The largest make the most vigorous shoots. These 
seem to be wholly devoted to the increase of the woody 
system of the tree. Lower dow^n, the strong last year's 
shoots arise from buds not quite so large. These make 
shoots less vigorous than the other class, and bear the female 
flowers on their apices. Below these are numerous small 
weak buds, which either do not push into growth at all, or 
when they do, bear simply the male catkins." 

Again, Castanea Americana bears two crops of male flow^ers, 
the first of which disarticulate and are useless ; the second 
appear about ten days later, accompanied by clusters of 
females. Occasionally a tree will be entirely female. 

Mr. Meehan also calls attention to the fact that isolated 
trees of Bii'ch, though producing an abundance of male and 
female flowers, very often have not a perfect seed. Hazels 
are sometimes protogynous, sometimes protandrous ; and if 
the latter condition prevail, there may be little or no fruit, 
as often occurs in Pennsylvania. After making analogous 
observations on American Maples, he summarizes his remarks 
on the latter as follows : — 

" Male flowers do not appear on a female Maple-tree till 
some of its vital power has been exhausted. 

* Laws of Sex in J. Nigral Proc. Acad. N. Sci. of Phil., 1873, 
p. 290. 


" Branch-buds bearing female flowers have vital power 
suflBcient to develop into branches. 

"Branch-buds bearing male flowers have not vital power 
enough to develop into branches, but remain as spurs, which 
ever after produce male flowers only. 

" Buds producing male flowers only, are more excited by 
a slight rise of temperature than females, and expand at a 
low temperature under which the females remain quiescent" 
[i.e. when the winter temperature begins to give way to the 
rise in early spring, the males are more easily excited into 
maturity]. * 

As another authority, I would refer to a paper by Mr. 
Moore, upon the appearance of male flowers on female trees, 
such as the Papaw, etc He alludes to Dr. Wight's views, 
in that he attributes these changes " to the modifying power 
of the soil and climate acting on the dormant energies of the 
rudimentary ovaries and developing them into prolific fruit, 
but at the cost of the male organs." In another case of the 
Papaw one fertile flower was produced, and that the first 
which expanded, others being all male " It wonld seem 
that fertile flowers in these instances have only been de- 
veloped when the greatest vital energy is present in the 
plant, which is the case when they first begin to expand. 
Other instances," Mr. Moore adds, " might be quoted to show 
that vigour and healthiness increase the female line of vital 
force in vegetables, whilst weakness is more conducive to 
the male development." 

This view was corroborated by a case of a young plant 
of Nepenthes distillatoria, raised from seed. Mr. Moore 
describes and figures it in the same paper, The lowermost 
flowers of the raceme bore both stamens and pistil, the 

* On the Relation of Heat to the Sexes of Flowers, Proc Acad. Nat. 
Sci. of Phil., 1882, p. 1. 


carpels of which were somewhat dissociated. On the upper 
half they were entirely male. He did not succeed in impreg- 
nating any of the numerous and well-formed ovules. He 
observes : " This well -authenticated case also favours the 
theory that vigour in the plant is productive of the female 
line of vital force." * 

It is a common phenomenon for diclinous trees to change 
their sex in different places or seasons. Ashes and Maples, 
as well as Palms, have been known to do this. The only in- 
terpretation being apparently the difference which occurs in 
the cKmatal conditions from year to year, or the modifications 
of temperature, soil, etc., consequent on different environing 

Sexuality and Temperature. — Temperature has a marked 
influence on the sexes. A relatively high temperature favours 
the corolla and androecium, while a comparatively lower 
one the gynoecium. A. Knight long ago found that Water- 
melons grown with a maximum of 110° by day, usually 
varying from 90° to 105°, with a minimum of 70° at night, 
grew with luxuriance, but bore no fruit, though it had a 
profusion of minute male blossoms. This experience is 
corroborated by present horticulturists. He was not sur- 
prised, as he had for many years previously succeeded, by 
long-continued. low temperature, in making cucumber plants 
produce female flowers only. 

J\lr. Median's observations on tha development of buds 
on certain trees appeared to corroborate this view of 
Knight's. He remarks that, in the year 1884, after a winter 
of uniformly low temperature, the male and female flowers 
of the nut appeared together ; but in other years it was 

* Trans. Irish Acad., xxiv., p. 629; see also a paper on " Sexuality," 
by Dr. M. T. Masters, Pop. Sci. Rev., xii., p. 363, 1873, and his Teratology, 
p. 190 J also, Proc. Acad. Nat. Sci. of Phil., 1873, p. 290. 


found that a few warm days in winter would advance the 
male flowers, so that they would mature some weeks before 
the female flowers opened. Hence the latter were generally 

That the stamens are much more sensitive to and pre- 
cocious in their development under a rise of temperature, is 
seen in the behaviour of plants in different countries. Thus 
it is asserted f that Stratiotes aloides produces its carpels 
with greater abundance towards the northern limit of its 
geographical distribution, and its stamens, on the contrary, 
are more frequently developed in more southern districts. J 

These tendencies to check one or the other sex, may lead 
to monoecious diclinism; and even complete dicecism seems, 
at all events to some extent, due to climate, as differences 
occur in widely separated countries ; thus Honclienya peploides 
is frequently hermaphrodite in America, but usually sub- 
dicecious in England. § 

Mr. Darwin, in his experiments, found that Mimulus 
luteus was very sterile in one year ; and he attributed the 
fact partly to the extreme heat of the season. || 

* Proc. Acad. Nat 8ci. of Phil, 188 i, p. 116. 

t Teratologrj, p. 196. 

X Perhaps the propagation by apogatny of the female plants of 
Chara crinita may be a resonrce to which this plant has been driven in 
consequence of the male plants not thriving in a cool region. Sachs 
says that the female is found throughout the whole of Northern Europe, 
but the male is only known to occur in Transylvania, South of France, 
and by the Caspian (Phys. of Plants, p. 801). 

The idea is suggested by this that when temperature arrests the 
male without checking the vegetative system, a plant may adopt 
vegetative methods of multiplication. Thus, instead of regarding the 
" root-tubers " and aerial corms of Ranunculus Ficaria as the cause of 
the degeneracy of the pollen in that plant ; perhaps it would be more 
correct to reverse the process. 

§ Teratology, p. Id6. \\ Cross and Self Pert., etc, -p. Q8. . 


Mr. Darwin also records * how " a tendency to the 
separation of the sexes in the cultivated Strawberry seems 
to be much more strongly marked in the United States than 
in Europe; and this appears to be the result of the direct 
action of climate on the reproductive organs." Quoting 
from the Gardener s Chronicle, f he adds, " Many of the 
varieties in the United States consist of three forms, namely, 
females, which produce a heavy crop of fruit, — of hermaphro- 
dites, which ' seldom produce other than a very scanty crop 
of inferior and imperfect berries,' — and of males which pro- 
dace none. . . . The males bear large, the hermaphrodites 
mid-sized, and the females small flowers. The latter plants 
produce few runners, whilst the two other forms produce 
many; ... we may therefore infer that much more vital 
force is expended in the production of ovules and fruit than 
in the production of pollen." 

Conversely, as runners were more abundant with male 
and hermaphrodite plants, we see here an instance of vege- 
tative growth correlated with the male elements at the expense 
of the female. 

Sexuality and the Soil. — Miiller has given two instruc- 
tive cases where it is pretty certain that the soil was a chief 
cause of the separation of the sexes. J Dianthus deltoides, 
near Lippstadt, offers interesting gradations from her- 
maphroditism to gynodioecism and gynomonoecism. " On 
the border of a meadow, of some hundred stems examined by 
myself, all the flowers, without exception, proved to be pro- 
tandrous, with a normal development of the anthers and 
stigmas. On the grass-grown slope of a sandy hill likewise, 
all the stems produced protandrous flowers, but on many 
stems the stamens, although emerging above the petals 

* Farms of Flowers, p. 293. f 1861, p. 716. 

X Nature, vol. xxiv., p. 532. 


before the development of the styles and stigmas, hore 
diminished whitish anthers, not opening at all, and containing 
also some shrivelled pollen-grains. Lastly, in a barren sandy 
locality, many of the stems produced female flowers, Tvith 
stamens aborted in the same degree as in D. superhus, and 
not infrequently such female flowers and protandrous her- 
maphrodite ones are found on the same stem." Wiegman also 
found the Diantlius had contabescent stamens when growing 
on a dry and sterile bank. The conditions here mentioned 
are very like those more than once described as associated 
with double flowers, in which the stamens have also de- 
generated but taken the petaloid form. Hence I think we 
may directly trace the degeneracy of the anthers and pollen 
to atrophy ; since chemical analyses of pollen prove that the 
most important constituents required are potash, nitrogen, 
and phosphorus pentoxide,* probably wanting in the localities 

" Centaurca Jacea " Miiller describes f " as having its 
flower-heads of the same stem ahvays of the same form, but 
different stems of the same locality often present astonishing 
differences in their flower-heads. 

"In the most common and apparently original form, the 
flower-heads consist of florets which are all of the same 
tubular shape, and all contain both fully developed anthers 
and stigma, the divergence of tlie outer florets giving to the 
whole head a diameter of 20-30 mm. From this original 
form variation has gone on in two opposite directions, the 
final effects of this variation being, on the one side, very 
conspicuous male flower-heads of 50-55 mm. diameter ; and 
on the other side less conspicuous female flower-heads of 

♦ From an analysis of Ash blossoms, by Professor Church, Journal 
of Botany, 1877, p. 364. 

f Nature, vol. xxv., p. 241. 


30-35 mm. diameter. In both these extreme forms the outer 
row of florets possesses greatly enlarged radiating corollas 
which are sexually functionless, but useful in making the 
flower-mass more conspicuous. In the male flower-heads, 
anthers and pistils of the disk-florets are well-developed, 
but the style-branches never open so as to expose their 
stigmatic surfaces, and in their basal portion are grown 
together. In the female flower-heads, on the contrary, only 
the pistil of the disk-florets is fully developed, the anthers 
being pollenless, shrivelled, and brownish coloured. 

" These two extreme forms are linked with the original 
one by a continuous series of gradations When in the 
original form variation begins in one direction, the outer 
row of florets gradually becomes longer and more i^adiating, 
and in the same degree their sexual organs diminish in size 
and become functionless, the anthers first aborting, and then 
the pistil. Finally, the barren ray-florets continuing to 
increase, the pistils of the disk- florets, too, become function- 
less, and the conspicuous male flower-head is accomplished. 

" In the contrary variation some of the outer florets of 
the original form begin to diminish in size, while their 
anthers become brownish and pollenless, and this change 
step by step proceeds inwards and seizes a greater and 
greater number of disk-florets, until the whole flower-head 
is female, and reduced to a diameter of 15-18 mm. This 
state being reached, the corollas of the marginal flowers 
recommence to increase and become radiating, while at 
the same time their anthers disappear without leaving any 
trace, and their style-branches remain closed together." 

Calendula officinalis furnishes another instance of com- 
plete change of sex, most probably caused by varying con- 
ditions of nutrition supplied by the soil. In the normal 
" single " form the disk florets are male, but with club- 


shaped stigmas. The two style arms, being fused together 
and strongly papillose, are only useful for thrusting out the 
pollen fi^om the anther cylinder. In " double " forms the 
corollas all become ligulate, the stamens disappear altogether, 
and the style arms of the pistils assume the normal form 
characteristic of the ray florets. They now set seed, so that 
the entire capitulum is female, and forms fruit.* 

Polygamous states often occur in trees growing apparently 
under the same conditions, and although we cannot doubt 
that they are due to different degrees of nutrition, yet they 
cannot be readily correlated to ^dsible differences in the 
environment, Mr, Darwin thus describes the Ash : f "I 
examined fifteen trees growing in the same field ; of these, 
eight produced male flowers alone, and in the autumn not 
a single seed; four produced only female flowers, which set 
an abundance of seeds , three were hermaphrodites, and two 
of them produced nearly as many seeds as the female trees, 
whilst the third produced none, so that it was in function 
a male. The separation of the sexes, however, is not com- 
plete in the Ash; for the female flowers include stamens, 
w^hich drop off at an early period, and their anthers, Avhich 
never open or dehisce, generally contain pulpy matter instead 
of pollen. On some female trees, however, I found a few 
anthers containing pollen-grains apparently sound On the 
male trees most of the flowers include pistils, but these 
likewise drop off at an early period , and the ovules, which 
ultimately abort, are very small compared with those in female 
flowers of the same age." 

It may be added that the staniens are sometimes sub- 

* I found no difference whatever between the plants raised from 
the larger seeds of the ray florets and the smaller ones of the disk 
florets. They all gave rise to the " single " form of capitulum. 

t Forms of Flowers, p. 11. 


petaloid forming staminodia — another hint that " conta- 
bescence " is closely akin to petalody of the androecium. 

Sexuality and Heterogamy. — Another source of diclinism 
may theoretically be attributed to protandry and protogyny 
carried to such a degree that the opposite sex is arrested 
altogether. Many plants have their flowers hovering about 
homogamy, some individuals being protandrous, others proto- 
gynous, according to locality, etc. Thus Saxifrages and 
species of Ribes are in this condition. 

We know that as soon as a flower is fertilised, the corolla 
fades and mostly falls. This means that the nourishment is 
now directed into the pistil. In a protogynous flower the 
petals and stamens may be in a very undeveloped state, while 
the stigma is ready for pollination.* If it be fertilised it 
no longer requires other organs, and nourishment may be 
abstracted from the corolla and stamens, which therefore 
would tend to abort. Let this procedure become hereditary^ 
and we get passages to female flowers. Moreover, the more 
female forms tend less to degeneracy, plant for plant, than 
the hermaphrodites, as Darwin showed with Satureia, and as 
is known to be the case with Strawberries in the United 
States, and again as is the case with the Ash, described above. 
Therefore female plants might be produced abundantly which 
would keep that form permanent. 

Conversely, plants growing in the open with an increase 
of temperature, and readily seen and visited by insects, 
become strongly protandrous ; consequently the pistil is at 
first delayed in development with a corresponding tendency 
to enfeeblement in comparison with the more purely female 

The results of crossing these conspicuous flowers — and 

* See e.g. Miiller's figures of Saxifraga Seguieri in different stages, 
Fertilisation, etc., p. 244. 


the more conspicuous the more masculine is the flower, 
and the more attractive will it be — one with another, would 
not therefore be so advantageous as crossing the more female 
plants with the conspicuous. The former, too, produce 
relatively more offspring, and might tend to oust the others, 
and reproduce both the "more masculine" and the "more 
female " sorts. Intercrossing, therefore, coupled with en- 
vironing conditions, may together bring about dioecism, as 
in Strawberries. As this reasoning is rather deductive^ it 
must be only considered as a suggestion. 

Sexuality axd Heterostylism. — This undoubtedly is 
another source of diclinism, as already alluded to. Mr. 
Darwin alludes * to Coprosma and Mitchella as indicating 
this fact. " Coprosma is dioecious, and in the male flowers 
the stamens are exserted, and in the female flowers the 
stigmas-, so that, judging from the affinities of these genera, 
it seems probable that an ancient short-styled form, bearing 
long stamens with large anthers and large pollen-grains (as 
in the case of several Rubiaceous genera), has been converted 
into the male Coprosma; and that an ancient long-styled 
form, with short stamens, small anthers, and small pollen- 
grains, has been converted into the female form. According 
to Mr. Meehan,t Mitchella repens is dioecious in some 
districts t for he says that one form has small sessile anthers 
without a trace of pollen, the pistil being perfect ; while in 
another form the stamens are perfect and the pistil rudi- 
mentary. Mitchella, therefore, would seem to be heterostyled 
in one district and dioecious in another," and this can 
scarcely be due to anything but environment. 

* Forms of Flowers, etc., p. 285. See also above, p. 228. 

t Proc. Acad, of Sci. of Philadelphia, July 28, 1868, p. 183. I do not 
gather from Mr. Meehan's account that he found any difference as to 
locality. Dioecism appears to be a constant character. 


Summarizing the various influences of the environment 
as climatic — such as temperature and light, shade and 
obscurity, humidity and drought, a^; well as varieties of soil 
and degrees of nourishment, and possibly others — we soon 
see how careful one must be in attributing a result to any 
one or special cause alone. What we can do is, as it were, 
to pick out of them, as tolerably well-ascertained, condi- 
tions which seem to favour, say, the female as compared with 
the male organs or flowers — such as, e.g., a mean or optimum 
condition of vegetative energy, a relatively low temperature, 
no excess of nutriment, a due amount of light, humidity, 
etc.; or again, on the other hand, a relatively higher tempera- 
ture, which favours and stimulates the staminal energies, the 
androecium being more keenly sensitive and more readily 
responsive to slight increments of temperature than is the 
gynoecium. The duration of the male elements being shorter 
than that of the female, they can come more quickly to 
maturity and perish earlier, as seen, for example, in the first 
flowering deciduous male catkins of Castanea Americana 
mentioned above. These, having been formed at the close 
of the preceding year (like many male flowers of the 
Umhelliferce late in the season), may represent the ex- 
piring energy of the year's growth. They open first, as 
soon as a sufficient though slight increment of temperature 
occurs, but quickly fall off, quite useless, as no female 
flowers are open to be benefited by them. 

Again, many, if not the majority of gjnodioecious plants 
would seem to be produced by the first flowers opening 
before the temperature was sufficiently high to allow of the 
corolla and stamens to develop properly ; and though many 
female flowers of the Lahiatce now blossom simultaneously 
with the hermaphrodite flowers of the same species ; this 
may be, perhaps, accounted for by hereditary influences, as 


Mr. Darwin showed that seeds of the female plants of Thyme 
yielded both female and hermaphrodite plants. 

Although, therefore,* we are nnable to fathom all the 
mysteries of Nature's proeedure, we can detect some of the 
lines upon which she works, and perceive how, in all cases, 
it is the environment — but sometimes one set of influences, 
sometimes another — which, being brought to bear upon the 
plant, the latter responds to it; and some form of what 
may be called " incipient diclinism " is the first result. If, 
then, these influences be kept up, hereditary conservatism 
comes into play, and such slight beginnings towards a 
separation of the sexes becomes fixed — only temporarily, 
however, — which constitute the first step, to be followed by 
others, till absolute and almost irrevocable dicecism is the 
final result. 

Dr. M. T. Masters has collected several cases in which 
one or other of the sexes has been arrested, apparently in 
consequence of the nature of the soil and other conditions of 
the environment. I refer the reader to his " Teratology," as 
my object is not merely to enumerate all the instances known, 
but suflicient to establish the theory advanced, — that it is the 
environment that first influences the organism, which then 
responds to it; and that, secondly, all adaptive variations 
thus set up — provided the environment continue to exert its 
influences — can become fixed by heredity. The consequence 
is that they are ultimately recognized as constant and 
specific characters. 

The Origin of Sex. — If now the environment has been 
proved to exert potent effects upon the development of the 
sexual apparatus of flowers, there still remains the ques- 
tion how far is either sex or both present, or at least poten- 
tial, in the embryo. Marked differences have resulted from 
sowing fresh or w^ell-matui'ed and older seeds of melons. 


M. Arbanmont found tliat young seeds gave rise to plants of 
extraordinary vegetative vigour ; moderately aged ones gave 
rise to corresponding moderately vigorous plants with both, 
male and female flowers ; while older seeds gave rise to still 
less vigorous plants, but which, when properly nourished, 
formed female buds.* M. F. Cazzuolaf also found that 
melons raised from fresh seed bore a larger proportion of 
male flowers than female ; while older seed bore more female 
flowers : and this has been confirmed. 

Another interesting result was obtained by M. Triewald, 
who grcAv twenty-one out of twenty-four melon seeds which 
were forty-one years old. The branches were very narrow, 
yet they produced early and plenty of good melons. [j: A 
cause of the differences of vigour in the plants raised from 
seeds of different age is, perhaps, connected with, the fact 
that fresh melon seeds contain a neutral oil, which becomes 
more and more acid by keeping. This increased acidity 
coincides with a diminished germinative power ; § and 
proportionately, therefore, less liable to run into excessive 
vegetative growth. 

The next condition to be observed is that resulting from 
sowing seeds of diclinous plants thickly or thinly. Hoff- 
man's experiments || in this direction showed that 283 male 

* Bull, de la Soc. de Bot. de Fr., 1878, p. 111. 

t Bull, de Tuscan. Hort. Soc, 1877. 

+ Gard. Chron., 1879, p. 470. 

§ M. Ladureau in Ann. Agronomiques. Mr. Darwin also found that 
fresh seeds of Iheris grew at first more vigorously than others {Cross 
and Self-fertilisation, etc., p. 103). 

II Gard. Chron., 1879, p. 762 ; see also Bot. Zeit, xliii., 1885, p. 145, 
seqq. ; also Jenaisch Zeitschr. f. Naturwiss, xix. (1885), sup. ii., pp. 108- 
112. The following were the plants with which he experimented : 
Lychnis diurna, L. vespertina, Valeriana dioica, Mercurialis annua, Buniex 
Acetosella, Spinacia oleracea, and Cannahis sativa. 


plants appeared, and 700 female, in the thickly sown plot, 
while only 70 males occurred when thinly sown. This has 
been paralleled in America, where Mr. Meehan, of Phila- 
delphia, has noticed how Amhrosia arte misisef alia, if growing 
vigorously, has a proportion of female flowers largely in 
excess of the males ; but in fields where the grain has been 
cut, and this " Rag- weed " comes up in thick masses late in 
the season, the individual plants nearly starving each other, 
male flowers are very numerous, and some are wholly male. 
Prantl also observed that the crowded prothallia of Ferns gave 
rise to more antheridia, and scattered ones more pistillidia. 
Pfeffer, too, noticed the same fact with Eqitiseium. 

. In these cases w^e seem to have results exactly the reverse 
of those of the melon seeds : but while in the latter the male 
flowers were accompanied by the precocious and excessive 
vegetative energy, the female were prevented from appearing 
at all ; for it must be remembered that normally male 
flowers of melons appear before the females. In the case 
of thin sowing, the . plants were in a natural and healthy 
condition : but when crowded they were starved, and the 
vital energy, being just enough to develop male flowers, 
proved insufficient for the female ; and, conversely, when 
thinly soTvm, "vitality" was not checked, and females were 

The question arises, are all seeds potentially bisexual, 
and one sex rather than another determined either by an 
inherent vigorous constitution or by the conditions of the 
environment during germination and growth ? or is there, so 
to say, a determination of sex, or at least a predisposition, 
at an earlier stage still ? Dr. Hoffman, judging from his 
experiments, is inclined to the opinion that sex does not 
reside in the seed, but depends on conditions of germination. 
Mr. W. G. Smith arrived at the same conclusion, for he says 


in his ternaries on some Bicecious Plants,^ " 1 think seeds 
themselves are probably not either male or female, but that 
after influences produce the sex ; as in animals the sex is not 
developed in the early embryo life of the creature, nor till 
the embryo has attained a certain age." 

On the other hand, F. Heyer thought sex was determined 
at an earlier period than the rijDening of the seed.f Some 
differences which have been noticed in seedlings of Nutmegs 
seem to countenance this idea; thus Mr. Prestoe, in his 
report on the Trinidad garden, ;{: says that " the leaf of the 
female seedling is most perfectly elliptical, with straighter 
primary veins. In the male plant it is broader towards the 
point than at the middle, i.e. obovate, and furnished with a 
point much longer than that of the female. The veins are 
also curved in towards the point much more roundly than 
in the latter." 

An interesting experiment by Mr. T. Anderson-Henry, 
recorded in the Gardeners Chronicle of 1876, may be quoted. 
He says, " I raised a seedling Begonia having female flowei'S 
only. It resulted from an experiment I made on the seed- 
bearer by cutting off two of the three lobes which compose 
the stigma, and fertilising the remaining lobe, I repeated 
this experiment ; and all of the progeny which have yet 
bloomed, consisting of four or five plants, have likewise all 
come with female flowers only." This seems to show that 
the female seedlings were due to concentration of energy to 
a limited number of seeds. On the other hand, a hybrid 
Begonia^ " Adonis," raised by Mr. Veitch from a summer- 
flowering tuberous variety, " John Heal," crossed with a 
■winter-flowering variety (itself obtained from B. Socotrina 
crossed by a dwarf- flowering tuberous variety), bore nothing 

* Journ. ofBot, 1864, p. 232 (note), t Journ. Micr. Soc, 1884, 251. 
X Gard. Chron., 1884, p. 315. 


but male flowers — presumably in consequence of some weak- 
ness of constitution due to hybridisation. 

It would be quite foreign to my purpose to trace the 
origin of sexes throughout the vegetable kingdom, as I am 
solely concerned with that of flowers. But what appears to 
be pretty certain is that the absorption of the pollen-nucleus 
by the " egg-cell " involves a special form of nutrition, 
coupled with certain excitant effects. Union between nuclei 
occurs elsewhere ; and as illustrative analogies, one recalls the 
fact of fusion being normal in the Conjugates^ and among 
zoospores, where no sexual differentiations are observable. 
Again, in the embryo-sac there occurs the union of two nuclei, 
one from each tetrad, their function being then apparently to 
form endosperm. As another case, Mr. Gilburt has described 
the union of the nuclei of cells constituting a "cell-group," 
which forms a wood-fibre after the absorjDtion of the septa.* 

Of course one of the most essential properties of the 
pollen-nucleus is to transmit to the offspring characteristics 
of the male parent : but even this is paralleled in the vegeta- 
tive system ; for an engrafted scion can transfer its peculi- 
arities to the stock, as has occurred with Cytisus Adami, 
variegated Abutilons, etc. 

If, however, we. ask what are the actual differences which 
exist between the male and female energies, and how they 
have arisen, we at once find that we are completely bafiSed, 
and that all speculations are at present futile. 

* Morph. of Veg. Tiss., Journ. Roy. Micr. Soc, 1879, p. 806 (note). 
Schacht observed a similar origin of liber-fibres in the Papaw, each of 
which was originally composed of three or four cells, but the septa 
become absorbed; their original positions being only indicated by 
clusters of pores on the walls {Les Laticif. du Carica Papaya, Ann. des 
Sci. Nat., 4 ser., viii., pi. 8, figs. 9, 10). Treub, on the other hand, dis- 
covered the laticiferous vessels and liber- fibres of the Nettle, etc., to have 
arisen by repeated division of the nucleus, the partitions not having been 
formed at all (Arch. Neerl. des Sci. Exac. et Nat, torn, xv., 1880, p. 39). 



Inconspicuous and Oleistogamous * Flowers. — Degeneracy 
in plants is as of frequent occurrence as in animals ; and just 
as it implies no pathological or anything of a constitutionally 
injurious character in them, so, it must be distinctly borne 
in mind, does it imply nothing of the sort in plants. The 
word means "down from the genus ; " like " degradation," it 
is only a " step downwards." It implies retrogressive or at 
least arrested conditions ; but a degraded flower often 
acquires new features, qualifying it for securing self-fertili- 
sation with a far greater certainty than was the case with its 
more conspicuously flowering ancestors. 

There are several causes which can bring about degrada- 
tions in the various organs of plants, such as growth in 
water, subterranean habits, parasitic and saprophytic states, 
freedom from strains, compensation, etc. Though it would 
be interesting to trace out the cause and effect in each case, 
I must content myself with flowers, and particularly the 
essential organs. 

There are two principal causes which may be styled the 
rationale of degradation in flowers. The first is compensa- 
tion, when the vegetative system is in too great activity to 

* Cleistogamous, "a closed nnioD," i.e. when flowers are self- 
fertilising without opening. 


allow of the proper amount of nutrition being at the service 
of the flowering process. This is so well known that I need 
not dwell upon it now. The second is the cessation of insect 
fertilisation. The effect of fertilisation operates in two 
directions. On the one hand, if it be the result of inter- 
crossing bj' insect agency, it stimulates the flowers till they 
become thoroughly adapted to their visitors, and highly 
differentiated in certain ways in consequence, but more 
especially as regards the perianth and stamens ; while, in 
many cases, some degree of degradation occurs simultaneously 
in the pistil. Conversely, self-fertilisation and anemophily, 
consequent ujDon the neglect of insects, are accompanied by 
corresponding degradations in the perianth, stamens, and 
pollen, correlated with a regained ascendancy in the powers 
of reproduction. The limits of degradation, with an increase 
of fertility, are seen in many cleistogamous flowers. 

In tracing the progress of degeneracy from a species with 
large flowers to one with inconspicuous blossoms, I do not 
mean to imply that we can actually witness the process in 
activity • but we can see this represented, as it were, in many 
a series of what we call species of a genus ; but which we 
might call transitional forms of one kind. It is only because 
we cannot trace the actual process going on that we regard 
them morphologically as distinct species. Thus, if a verifiable 
demonstration be unattainable, it is a "moral conviction,'* 
not only that Geranium pratense is as much and obviously 
adapted to insect agency as G. pusillum is to fertilise itself, 
but that the latter species has been derived from the former 
or from some kindred plant, through some such transitional 
forms as G. pyrenaicum and G. molle. 

This process of degradation from insect to self-fertilising 
conditions, not only affects the size of all parts of the flower, 
but the entire plant. Mr. Darwin showed how the stimu- 


lating effect of crossing generally increased the heights and 
weights and, for a time, the fertility of the plants experi- 
mented upon. Conversely, self-fertilised species are alto- 
gether smaller than their allied intercrossing species. Thus 
Stellaria Holostea may be compared with S. medium, Cerastium 
arvense with G. tetrandrum and G. glomeratmn, Gardamine pra- 
tensis with G. Idrsuta, Polygonum amphihuun with P. aviculare, 
etc. Besides being thus dwarfed, self-fertilising plants are 
mostly annuals. But while conspicuous flowering plants 
blossom during a limited period in summer only, their 
smaller, less conspicuous, and regularly self -fertilising allies 
may, and often do, flower and set seed all the year round. 

In my essay on " The Self- fertilisation of Plants,"* I 
drew up the following list of peculiarities of habitually self- 
fertilising plants, all of which indicate points of degeneration 
or arrest. 

1. The inconspicuousuess of the flowers, even when fully 

2. The calyx and corolla are often only partially expanded, 
or not at all. 

3. The white or pale colours of the corollas ; while 
specially coloured streaks, specks, "guides," and "path- 

* I must refer the reader to tlie above essay for a full discussion of 
this subject. The evidence there given proves conclusively that self- 
fertilising and anemophilous plants are in every way the most widely 
dispersed of flowering plants, and best fitted to maintain themselves 
in the struggle for life. I will add here that Mr. H. 0. Forbes came 
independently to a similar conclusion when studying cleistogamy in 
orchids ; and remarks, at the close of his paper {Journ. Lin. Soc, vol. xxi., 
BoT., p. 548), " The observations above given would seem, therefore, to 
support the Rev. G. Henslow's conclusions so ably given in his ' Memoir 
on the Self -fertilisation of Plants,' already published in the Transactions 
of the Linnean Society. My absence abroad prevented my seeing this 
paper till quite recently, and after I had completed these notes." 


finders " peculiar to intercrossed flowers are more or less 
reduced, if not absent. 

4. The partial or total arrest of the corolla. 

5. The mature stamens of the expanded flower retain in 
many cases the incurved, i.e. an arrested position, which 
they had in bud ; the anthers thus remain in contact with 
the stigmas. 

6. The stamens are often reduced in size and number, 
and the pollen in quantity. 

7. The pollen tubes may often be seen to be penetrating 
the stigmas, either from grains still within the anther-cells, or 
evidently derived from those of the same flower. 

8. The styles are shortened, and the stigmas are situated 
appropriately for direct pollination from the anthers of the 
same flower. 

9. The partial arrest of the corolla and stamens in their 
rates of development, allows the pistil to mature with com- 
parative rapidity. 

10. The consequent early maturation of the stigma, so as 
to be ready before or simultaneously with the dehiscence of 
the anthers. 

11. Little or no scent, 

12. Decrease in size or total absence of honey glands, 
with corresponding little or no secretion of honey.* 

Notwithstanding these various indications of degradation, 
such flowers are often con^elated with special alterations 
which secure self-fertilisation without a cbance of failure 
— a precariousness which almost always exists in flowers 
adapted to insects. Thus — contrary to the old but erroneous 

* Miiller, in his ''General Retrospect" (Fertilisation, etc., p. 591), 
also gives a number of modifications, mostly referred to in the text above, 
of what he describes as " the countless ways in which plants revert to 
self-fertilisation in default of sufiScient insect visitors." 



dictum that, whether flowers were pendulons or erect, the 
stigma was always heloiv the anthers, so that pollen could 
fall upon it — the anthers are always closely applied to the 
stigmas, as may be seen in Chickweed (Fig. 52), and small- 

Fig. 52. — riosver-bud, closed and expanded, 
of Stdlaria media, showing petals reduced 
in size; stamens, three only; anthers 
closely adpressed on stigmas. 

e 7> 

Fig. 53. — Stamens and stigmas of 
Epilobium montanum, the bud 
scarcely open, while anthers are 
closely applied to the stigmas. 

flowered Willow Herbs (Fig. 53), and especially in cleistoga- 
mous flowers (Figs. 56-59, pp. 258-261). 

The structure of the anthers and stigmas is often 
greatly altered in form, besides being merely reduced in 

As an illustration of the above remarks, the genus Viola 
is interesting as furnishing two " forms " of the same species, 
V. tricolor, or Pansy, the one 
being adapted to insects, the other 
to self-fertilisation ; while other 
species, such as F, odorata, the 
Yiolet, bear cleistogamous buds 
on the same plant as the ordinary 
violet blossom. 

The dimorphic flowers of Viola 
tricolor were first noticed by 
Miiller, who described them as follows : * "In the large- 
flowered form, the stigmatic cavity (Fig. 54, a, st) lies some- 
* Nature, Nov. 20, 1873, p. 45. 

Fig. 54.— Styles and stigmas of the two 
forms of Pansy • a, that of the larger 
and intercrossing ; b, that of the self- 
fertilising form. 

2dQ the structure of flowers. 

what more towards the top of the skull-like end of the style 
than in the small-flowered, one (6). When the skull-like knob 
in the two forms is pressed against the lower petal, in the 
large-flowered form the opening of the stigmatic cavity is 
directed outwards, so that the pollen-grains which have 
fallen out of the anther-cone can never spontaneously fall 
into the stigmatic cavity, and must be carried there bj 
insects ,- whereas in the small-flowered form the opening of 
the stigmatic cavity is directed inwards, so that pollen-grains 
falling out of the anther-cone spontaneously, fall directly 
into the stigmatic cavity. 

" In the large-flowered form, the opening of the stigmatic 
cavity (st) bears, on its lower side, a labiate appendage (I) 
provided with stigmatic papillas, so that a proboscis inserted 
into the flower when charged with pollen from a previously 
visited flo\ver, rubs off this pollen on to the stigmatic lip, 
thus regularly effecting cross-fertilisation ; whereas, when 
withdrawn out of the flower, charged with pollen, the 
proboscis presses the lip (Z) against the stigmatic opening 
(st), thus preventing self -fertilisation. This nice adaptation 
to those visitors provided with a long proboscis (Lepidoptera, 
ApidaD, Rhingia) is completely wanting in the small-flowered 
form (h). 

*' In the large-flowered form, there is a black wedge-shaped 
streak (g) on the front of the style, to which Mr. A. W. 
Bennett first called attention, and which he has interpreted 
as a guide-mark for those visitors which are diminutive 
enougli to crawl entirely into the flower. This streak is 
also wanting iu the small-flowered form. 

" In the large-flowered form, pollen-grains do not spon- 
taneously fall out of the anther-cone before the flower has 
been fully developed for several days ; whereas, in the small- 
flowered form, in bb far the majority of cases, a great number 


of pollen-grains fall spontaneously oat of the anther-cone 
into the stigmatic cavity and there develop long pollen- 
tubes, even before the opening of the flower, in much rarer 
cases a short time after it has opened. 

" When the visits of insects are prevented by a fine net, 
the flowers of the small-flowered form wither two or three 
days after opening, every one setting a vigorous seed- 
capsule ; those of the large-flowered form remain in full 
freshness more than two or three weeks, at length withering 
without having set any capsule ; when fertilised they, too, 
wither also after two or three days." 

I have met with several variations in minor details of 
structure in the smaller-flowered kind. Thus in some the 
stigmatic lip, probably representing one of the three stigmas, 
formed a globular knob 
protruding from the j Q^a 
orifice, as shown in Fig. ^-' "^ 
55, a, h. In another, 
it protruded like a 

tongue, C. The lateral Fig. 55.— styles and stigmas of self-fertilising forms 
T . 1 1 of Pansy. (For description, see text.) 

frmges,* which help to 

keejD the pollen back from reaching the stigmatic chamber in 
the larger flowers, are more or less retained in these ; as is 
also the bent-base to the style which forms the spring,* 
which keeps the globular head in a downward position. 

The accompanying figures will illustrate the cleistogamous 
flower-buds of Violets. They are very minute, about one- 
eighth of an inch in length (Fig. 56, /). The petals are 
reduced to linear and pointed structures, green or purplish 
green, (a) ; or they may be altogether wanting. The spur 
alone of the larger petal is sometimes present in strong- 

* For the theoretical origin of "fringes" and " springs," see Chap. 
XV., p. 133, and Chap. XIII., p. 123, respectively. 



growing garden plants (b). The stamens are five or less in 
number, having spoon-shaped connectives, and not pointed 
as in the normal form, bearing very minute oval anther-cells 
at the base (c, g) * Small bundles of pollen-tubes may be 
traced from the anthers into the stigma (g). The pistil has 
a short curved style, and truncated stigma (d) concealed 
beneath the anthers which lie imbricated over the top of 
the pistil. The anthers are usually devoid of appendages, 
though they are sometimes present, like the spur ; though 

Fig. 56.— Cleistogamous Violets, (For description, see text.) 

both organs are now useless. As the ovary swells it raises 
the stamens up with it (e). The capsules of the violet, Mr. 
Darwin observes, bury themselves in the soil, if it be loose 
enough, and there ripen ; but they certainly are very, if not 
more frequently not buried at all, but only concealed beneath 
the foliage. 

As another interesting case of a plant showing transi- 
tional conditions may be mentioned Scrophularia arguta, 
Ait.f " The two lowermost opposite and axillary branches 
bend backwards and penetrate the soil. The next pair do 

* (c) V, odorata; (g) V. canina. t Bull. Soc. Bot. de Fr., iii., p. 569. 


the same, but do not always reacli tlie ground, or else pene- 
trate it very slightly. They all bear fertile flowers. The 
lowest are apetalous, if completely hypogean [and pre- 
sumably cleistogamous]. Those which just reacli the soil 
have a corolla of four lobes nearly equal, and resemble the 
corolla of Vero7iica. A little higher up, the irregularity of 
the bilabiate character of Scropliularia is pronounced." 

The preceding quotation is interesting, first in showing 
how the subterranean cleistogamous form is derived from 
the conspicuous flower, and also supplies a hint as to the 
origin of Veronica^ in that it is a 4-merous degradation from 
a primitive 5-merous genus, which is lost or unrecognizable 
now, unless it be some member of the subgenus Pygmcea, 
which has five parts to the corolla.* 

As an illustration where geographical conditions favour 
the development of autogamous forms of flowers, the follow- 
ing passage may be quoted : — 

"Herr C. A. M. Lindman has examined the very rich 
flora of the Dovrefjeld in reference to the arrangements for 
fertilisation. He finds a distinct tendency to a deeper colour 
in the flowers than is displayed by the same species in the 
lowlands, red and blue predominating. The great length of 
daylight appears to increase the size both of leaves and of 
flowers, though, in some species, on the other hand, the 
flowers are diminutive in consequence of the low tempera- 
ture. Crowded masses of small flowers are very common. 
The number of scented species is comparatively small, 
though, the fragrance is sometimes powerful. The scarcity 
of insects necessitates that there should almost always be 
a provision for possible self-fertilisation ; and many species, 
elsewhere heterogamous, are here bomogamous. Notwith- 

* For Miiller's theory of the origin of Veronica, see Fertilisation, etc., 
p. 465. 


standing the cold and wet summer (1886), the plants 
observed almost invariably bore fruit."* 

As an example of pure cleistogamy I will take Oxalis 
Acetosella, as having special peculiarities. Mr. Darwin 
alludes to M. Michalet's description of the cleistogamous 
flowers of this species,t and adds some observations of his 
own.J He quotes an observation of Michalet's, that the 
five shorter stamens are sometimes quite aborted. This fact, 

which I have also ob- 
served (Fig. 57, d), is 
quite in keeping with 
S^ il| jW the common process of 

e d^^ the reduction of the 

Fig. 57.— Cleistogamous flower-buds of Oxalis number or parts of sta- 

Acetosella. (For description, see text.) . -.n o ,•!• • 

^ mens m selt-tertilismg 

flowers. He also adds this interesting observation : " In one 
case the tubes, which ended in excessively fine points, were 
seen by me stretching upwards from the lower anthers towards 
the stigmas, which they had not as yet reached. My plants grew 
in pots, and long after the perfect flowers had withered they 
produced not only cleistogamic, but a few minute open flowers, 
which were in an intermediate condition between the two 
kinds." This last remark is quite in accordance with the 
true origin of these flowers, that they are in all cases degra- 
dations from the conspicuous forms normally characteristic 
of the species which produce them. 

Fig, 57, a, clearly shows that in Oxalis Acetosella the 
cleistogamous state is simply a flower-bud which has become 
adapted to self-fertilisation ; and the intermediate conditions 
alluded to by Mr. Darwin I should suspect were analogous to 

♦ Journ. Roy. Micr. Soc, 1887, p. 615, and note. See below, pp. 270, 271. 
t Bull. Soc. Bot. de Fr., vii. (I860), p. 465. 
X Forms of Flowers, p. 321. 



58. — a, 
fulva ; b, 

of Impatiens 
stamens (after 

the permanent forms of the flowers of 0. corniciilata, whicli I 
at first inferred, from the wide distribution of this species, 
must be habitually self -fertilising. From Fig. 57, a, it will be 
seen that the corolla just protrudes from the closed sepals, 
and always remains as a "cap," h. Of the ten anthers, five 
are often abortive or wanting, d; the fertile anthers are 
placed over the very short stigmas, and are bound together 
by fine threads. These appear to play some part, bat the 
nature of their function is obscure, c. 

Impatiens fulva and I. Noli-me-tangere have also cleisto- 
gamous flowers. Fig. 58, a, represents 
a bud, and h two metamorphosed sta- 

Lamium amplexicaule will furnish 
another example of cleistogamy. This Fig 
genus has usually flowers highly differen- 
tiated, and adapted to insect fertilisation. 
That the cleistogamous flowers of this, as of all other species, 
are degraded forms of the normal kind is obvious from the 
presence of the "lip," as Avell as by there being four and 
didynamous stamens. The style elongates very much, and 
under the pressure of the closed 
summit of the corolla becomes bent, 
so that the stigmas lie between the 
anther-cells, and thus readily become 
fertilised. Fig. 59, a, represents a 
flower-bud ; h, the corolla in section ; «: 5 ^ 

and C, the pistil removed. This Fig. 59.— a, Cleistogamous flower- 

J. . j» 1 • • f 1 . hni of Laviium amplexicaule ; b, 

condition 01 cleistogamy is found m vertical section of same; c, pistil. 

the earlier-flowering plants, so that it is probably a mere 
result of check through a colder temperature. 

Salvia clandestina may be compared with the last 
described, as it is a self-fertilising form of, perhaps, S. jpra- 


tensis. Fig. 60,* a, represents a corolla, whicL. is very small, 
but open ; b represents the two fertile stamens ; the anther- 
lobes instead of being 
horizontal are erect, and 
face each other. The 
, u stigmas curl back be- 

Fig. eO.-Salviaclandestina: a, corollas ■,b,s.nthers; ^^^een them, aud are rc- 
c, style and stigmas. markablj long, C. 

The Origin of Cleistogamy. — We are now in a position 
to trace the causes of cleistogamy. Cleistogamous flowers 
nearly always occur on plants otherwise, or at least their 
allied species are, adapted for intercrossing, and include four 
genera of anemophilous plants. The first cause or influence 
is the arrest of the reproductive energy in the conspicuous 
flowers, which often set no seed at all. 

Whatever the primary cause of that may be, a very 
common result in perennials is to increase the power of 
vegetative methods of multiplication, as in the case of many 
bulbous and tuberous plants. 

This, however, is not a special feature of the plants which 
bear cleistogamous flowers. It would seem, therefore, that 
the reproductive energy being checked in one form of flower, 
it, so to say, breaks out in another. But there are several 
influences at work, and a very obvious one is temperature ; 
for the same species may behave very differently in one 
country with a high mean annual temperature, from what it 
does in another with a lower one. Thus, Viola odorata does 
not produce cleistogamous flowers in one part of Liguria, 
where the conspicuous flowers are perfectly fertile ; while 
they are mostly barren in England. On the other hand, 
cleistogamous flowers are produced by Violets near Turin, 

* From a specimen growing at Kew. It is cleistogamous at Halle 
(see below, p. 263). 


and abundantly in all parts England. Viola nana bears 
normal flowers in its native home in India, but only cleisto- 
gamous ones in England. Viola palusiris bears only the 
larger flowers near Paris, wbich are perfectly fertile, but 
when it grows on mountains it bears cleistogamous flowers. 
Similarly Impatiens fulva bears both kinds of flowers in Eng- 
land, but the larger are usually barren. After midsummer, 
in its native home in the United States, these flowers will 
produce capsules. Salvia clandestina, when transplanted 
from Africa to Halle, bore only cleistogamous flowers for 
five years, according to Ascherson, who considered the plant 
to afford an example of continuous self-fertilisation. He, 
however, afterwards observed ordinary open flowers. It is 
a species particularly common on the Continent. 

Again, plants vary according to the season. Thus Mr. 
Darwin found that Vandellia nummularifolia bore no perfect 
flowers in one season ; so, too, Ononis columnce bore none in 
1867, yet it had both kinds in 1868, 

The time of the year also influences the production of 
cleistogamous flowers. Thus Ononis minutissima^ liarvijiora, 
and 0. columnce, according to Mr. Bentham, produce them 
early in the spring. Godetia Cavanillesii and Lamium am- 
plexicaule do the same ; while some bear a fresh crop in the 
autumn, as 0. columnce. 

Two cases are mentioned by Mr. Darwin in which the 
period is the reverse of the above. 

Viola Boxhurghiana bore abundance of cleistogamous, but 
no perfect flowers, in Mr. Darwin's hothouse; and it bears 
the perfect flowers in India " only during the cold season, 
and these are quite fertile. During the hot, and more 
especially during the rainy season, it bears an abundance 
of cleistogamous flowers." * 

* Forms, etc., p. 320. 


The other example is Buellia tuhei'osa, of which Mr. 
Darwin remarks, " It produces both open and cleistogamous 
flowers; the latter yield from 18 to 24, whilst the former 
onlj from 8 to 10 seeds : these two kinds of flowers are pro- 
duced simultaneously, whereas in several other members of 
the family the cleistogamous ones appear only during the hot 
season." From this one would infer that an excess of heat 
may be a cause of cleistogamy, just as too low a temperature 
appears to bring it about. 

I think it probable that other influences than tempera- 
ture may be brought to bear upon a plant ; which, indeed, 
we may see in our own Yiolets. The larger flowers of this 
species are 7iot produced ia the hottest time of the year, 
while the cleistogamous buds are only borne in the summer. 
On the other hand, the foliage is only developed fully, con- 
temporaneously with the dwarfing of the floral organs. 

Again, a poor soil has been noticed as associated with 
cleistogamy by Torrey and Gvaj, in the case of North 
American species of Heliantliemum. 

Temperature, however, seems to be the most important 
agent; thus, while the climate of South Italy can develop 
the perfect flowers and render them fertile, there cleistogamy 
is suppressed; here, in England, the climate is seemingly not 
sufficiently warm to do so, and the cleistogamous buds appear 
in compensation. The vegetative energy, however, comes 
to the fore during the summer, and perfect flowers are not 
produced simultaneously with it ; so that it is not until the 
vegetative period has ceased, and the materials are remade for 
their development, that larger flowers are again borne later in 
the year, as in November, as well as in the following spring. 

With regard to the anemophilous genera, Mr. Darwin 
mentions Hordeumy Cryptostachys, Leersia oryzoides, and Juncus 
hufonius in Russia. 


Now, the three genera of Grasses here mentioned are 
characteristic of warmer regions, and even tropical, Leersia 
oryzoides being the sole species of that genus which reaches 
Europe, where it becomes cleistogamous. Therefore climatal 
conditions maj, with some reasonable presumption, be sug- 
gested as the immediate cause in these cases. "With regard to 
Hordeum murumm, which is, perhaps, almost habitually cleis- 
togamous in this country, it may be an hereditary result 
issuing from a similar cause. This may also apply to Viola ; 
for not only are some species tropical, but all the genera 
most nearly allied to Viola are tropical also. This is analo- 
gous to what I have suggested as the origin of gynodioecism 
in Lahiatce, which it may be noticed has at least two genera 
with cleistogamous flowers in this country or Europe. Juncus 
hufonius, according to Batalin, is exclusively cleistogamous 
in Russia, hence the same cause suggests itself for this 
species ; for, according to Ascherson, at Halle it has ordinary 
open, lateral, hexandrous flowers in addition to terminal 
cleistogamous triandrous ones.* This seems to show that 
lessened vigour has also a hand in the process in this case : 
the mean temperature of Halle is probably higher; if so, it 
may cause the plant to bear the open flowers there. 

From the above-mentioned facts, it will be seen that 
there may be more than one cause to account for cleistogamy. 
Hence, it must be regarded as an inevitahle result whenever 
those influences are brought to bear upon the plant which 
are capable of producing it ; and there is every reason to 
believe that whatever effects are produced in plants by 
external stimuli, if the latter be permanently kept up they 
will become hereditary, and then will be recognized by 
systematists as specific or generic characters. 

Anemophilous, or Wind-fertilised Plants. — The general 

* Miiller, I.e., p. 561. 


characters prevailing in this group consist of elongated 
papillose or plumose stigmas, or else thej spread out into 
laminae {EupJwrhia). The filaments are usually slender and 
movable, with versatile anthers, bearing incoherent and 
often smooth pollen-grains. In some cases the filaments are 
elastic, and project the pollen outwards ; or the whole flower 
may oscillate on a slender pedicel or peduncle, as the catkins 
of the Amentiferce, the flowers of Bumex, etc. Long, slender 
filaments are seen in Grasses, Sedges, Rushes, Hemp and 
Hop, Plantains, Littorella, and Poterium. Nettles and their 
allies are remarkable for their elastic filaments, which 
materially aid in the dispersal of the pollen. 

On the other hand, Palms, Bulrushes, etc., have more or 
less rigidly fixed flowers and floral organs. 

There is little doubt but that all wind-fertilised angio- 
sperms are degradations from insect-fertilised flowers. This 
is obviously so when many of the allies of an anemophilous 
genus or species are constructed for insects. Thus, Miiller 
says that Thalictrum minus* is anemophilous, while T. 
flavum is visited by several species of insects. Poterium 
Sanguisorha is anemophilous; and Sanguisorha officinalis 
presumably was so formerly, but has reacquired an entomo- 
philous habit; the whole tribe Poteriece being, in fact, a 
degraded group which has descended ir om Potentillece. Plan- 
tains retain their corolla, but in a degraded form. Juncece 
are degraded Lilies ; while Cyperaceoe and Graminece among 
monocotyledons may be ranked with Amentiferce among 
dicotyledons, as representing orders which have retrograded 
very far from the entomophilous forms from which they 
were possibly and probably descended. 

* I do not know on what reason ; for the stigmas are not charac 
teristic of such flowers. On d priori grounds I should have inferred its 
being self-fertilising, as the anthers completely conceal the few and 
small carpels. 


What, then, have been the causes which have given rise 
to the features generally characteristic of anemophilous 
flowers ? In the first place, it must be remembered that 
such are far from absolute. Smooth and easily scattered 
pollen,* Miiller remarks, is the only positive character 
common to these plants. Mr. C. F. White, F.L.S., however, 
tells me that from his researches he very much distrusts the 
division so generally accepted between wind- and insect- 
borne pollens. It is his opinion that there is no pollen-grain 
so smooth but that the hairs on the limbs of a bee or fly can 
hold it. Moreover, no pollen, however massed together, can 
possibly be heavier than, say, a thistle seed and its down 
attached, which the wind can carry with perfect facility ; so 
that to draw any distinction on that score seems to me to 
be very far-fetched.f With respect to the pollen of Grasses, 
Mr. White observes that it is perhaps forgotten that, although 
smooth in water, when dry they are notably wrinkled into 
sharply angled and irregular shapes. 

Mr. Edge worth J has figured many forms of pollen of 
anemophilous genera, several of which show no signs of 
smoothness or rotundity, such as Alopecurus prate7isis, Car ex 
arenaria, and G. panica, which, like Juncus effusus, is oblong, 
with sharp edges, all of which are at right angles or nearly 
so. Again, TypJia latifoUa and Cupressus have octahedral 
pollen ; Areca Baueri, Geratozamia, Bheum, Mercurialis, Oak, 
etc., have more or less sharply pointed spindle-shaped grains. 

* See Mr. A. W. Bennett's paper, On the Form of Pollen-grains in 
Reference to the Fertilisation of Flowers, Brit. Assoc. Rep., 1874. 

t I would here allude to another d prio7-i assumption. It has been 
thought that the two pouches on the pollen of the Fir aid it in trans- 
portation; but unless they were filled with some gas lighter than air 
they only increase the weight of the grain. 

X Pollen, by Mr. M. Pakenham Edgeworth, F.L.S., 1877. 


In Corylus, Alnus, and Plantago media, they are- polygonal, 
while Beech has them deeply three-grooved, etc. 

Mr. Edgeworth, in fact, states that the different kinds of 
pollen of anemophilous plants " are by no means all globular, 
as Mr. Bennett asserts." 

He notices, however, tbat " the grasses and Cyperaceoe, 
and perhaps the Flantagineoe are without the sticky nature 
of the outer coat, which obtains through all other pollen 

With regard to the versatile condition of the anthers in 
grasses, and their consequent facility of oscillating on a point, 
this feature seems to be only the result of the extremely 
slender filament due to degradation ; * and not quite the 
same thing as the antero -posterior oscillation which the 
action of bees has set np in the connectives of Salvia, species 
of Calceolaria, and Curcuma Zerumhet.-f Remembering how 
the rigidity of the filaments of intercrossing flowers is corre- 
lated to the retention of some well-defined positions for the 
anthers, so that insects can be struck by them accurately, 
and be again struck on the same spot by the stigmas of other 
flowers, we see that when the stimulus due to intercrossing 
has been long withheld, the filaments have become slender, 
easily waved about by the wind, and versatility of the 

* Plantago media, which is visited, has motionless anthers ; but in 
the anemophilous species of Plantain they are versatile. 

t Mr. H. 0. Forbes has described and figured a very analogous case 
in this species of Curcuma of Sumatra. The two anthers project for- 
wards in contact, they are provided with terminal processes like horns. 
The style passes between them. When a bee enters the flower it 
depresses these horns with its head, and so forces the anthers down- 
wards on to its thorax. The anthers bring the style and stigma down 
also. In a similar way do some species of Salvia cause the style to be 
brought down from the hood (A Naturalist's Wanderings in the Eastern 
Archipelago, p. 247) . 


anthers has followed. Those wind-fertilised plants with stiff 
filaments have presumably not yet degraded to a similar state. 
With regard to the pistil, since of heterostyled plants 
the stigmatic papillae are larger and longer in the long-styled 
forms, we seem to get a hint as to the origin of the papillose 
and plamose characters of many wind-fertilised plants ; in 
that such may be due to compensatory processes on the loss 
of the corolla, honey-secreting organs, etc., which have thus 
favoured the development of the pistil generally, such deve- 
lopments becoming emphasized in certain directions. 

Protogyny or homogamy generally accompany anemo- 
phily.* Thus Miiiler mentions Thalictrum minus, Flantago, 
Luzula, Callitriche, Myriojphijllum, and many Grasses as 
being protogynous ; and a common characteristic feature of 
such flowers is frequently noticed by Miiller, viz., that they 
have all "long-lived stigmas." This seems clearly to point 
to a relatively increased amount of vigour in the develop- 
ment of that organ in protogynous flowers ; which becomes 
especially noticeable in their enhanced size, as seen in most 
anemophilous flowers. Foterium he regards as homogamous, 
as well as Rye and Wheat. These conditions all agree with 
the total suppression of the corolla, and may be regarded as 
signs of degradation : and I have elsewhere shown, when 
treating of emergence and development of the floral organs, 
how a compensatory process accompanies the formation of 
the corolja and stamens on the one hand, and of the pistil 
on the other ; so that when the former tend towards degra- 
dation, the pistil gains the ascendancy, and matures earlier. 

* Artemisia vulgaris seems to be protandrous. The style arms are 
provided with papillose rosettes in the central florets, but are very 
elongated, and terminate in points in the circumferential florets. In 
no case could I detect pollen-tubes in unopened florets, though the 
grains were shed. 


Hence, to find its stigmas enlarging under anemopliily is all 
in keeping with the above facts. 

The Origin of Anemophily. — With regard to the origin 
of anemophilous flowers, there is every reason to believe 
them to be due to the neglect or absence of insects : that as 
these have brought about brilliant colours or other kinds of 
conspicuousness, so their absence has allowed flowers to 
degenerate and become inconspicuous, the result being 
either self-fertilisation or anemophily. As two examples 
of districts which illustrate this fact, are the Galapagos 
Islands, visited by Mr. Darwin, and Greenland, the flora of 
which is described by M. Warming. 

The former observer, on landing, thought that there were 
few or no flowers, but, on stricter search, discovered many 
to be inconspicuous. A specimen before me of Solanum 
nigrum, which he brought from those islands, has flowers 
much smaller than our own native plant, and illustrates the 
wide dispersion of self-fertilising plants. M. Warming 
found Greenland, like the Galapagos Islands, to be poor in 
insects, and "the flow^ers display a corresponding increased 
tendency to autogamy. One hundred and thirty-eight species 
of anemophilous plants are also named by him, exclusive of 
Willows. The flowers appear to decrease in size with the 
increase of latitude ; and the brilliancy of colour certainly 
does not become greater." * 

This last observation does not agree with M. Plahault's 
observations ; f and possibly M. Warming is here intimating 
a wrong cause of degeneracy, which I should incline to regard 
as the absence of insect stimulation, with the cousequent 
tendency to inconspicuousness, anemophily, and autogamy. 

* Overs. K. Banske Yidensk. SelsTc., 1886, p. xxv. (quoted from Journ. 
Roy. Micr. Soc, 1887, p. 433). See also above, pp. 177 and 259. 
t Ann. des Sci. Nat, 6 ser., t. vii. (1877), 'et t. ix. (1879). 


Where, however, insects are abundant, whether in high 
latitudes or greater altitudes, as in the Alps, there two 
causes will be at work to enhance the brightness of flowers ; 
viz. insect stimulation and prolonged sunlight. For Sachs 
has shown that the ultra-violet and invisible rays are 
specially efficacious in the development of flowers; and as 
the foliage grows more vigorously with prolonged light so 
it is presumable that the flower-forming substances will be 
more abundant as well.* 

The genus Plantago, like Tlialictrum minus, Foterium, and 
others, well illustrates the change from an entomophilous to 
the anemophilous state. P. lanceolata has polymorphic flowers, 
and is visited by pollen- seeking insects, so that it can be 
fertilised either by insects or the wind. P. media illustrates 
transitions in point of structure, as the filaments are pink, 
the anthers motionless, and the pollen-grains aggregated, and 
it is regularly visited by Bomhus terrestris (Delpino). On the 
other hand, the slender filaments, versatile anthers, powdery 
pollen, and elongated protogynous style are features of other 
species indicating anemophily; while the presence of a 
degraded corolla shows its ancestors to have been ento- 
mophilous. P. media therefore illustrates, not a primitive 
antomophilous condition, but a return to it ; just as is the 
case with Sanguisorba officinalis and Salix Cajprea ; but these 
show no capacity of restoring the corolla, the attractive 
features having to be borne by the calyx, which is purplish 
in Sanguisorba, by the pink filaments of Plantago, and by the 
yellow anthers in the Sallow Willow. Plantago alpina is 
self-fertilising, as the stigma does not wither until after 
maturing the anthers. 

If we may speculate as to why some degraded flowers 

* See La Vegitation du Globe, par Grisebach, t. i., p. 155 (trad. fran. 
de Tchihatchef). 


have become regularly autogamous, while others are now 
anemophilous, it may be due to the fact that, if a flower has 
been entomophilous and even strongly protandrous, the first 
stage of degradation is to bring the essential organs to a 
homogamoiis state. If they stop there, and become autoga- 
mous as well, which is the nsual result, then the flower will 
remain persistently self -fertilising, as, e.g., Shepherd's-purse, 
Chickweed, Knot-grass, etc. 

If, however, the flower had been protogynous, such as 
early-flowering Hellebores, Primus communis or some Alpine 
species, with "long-lived stigmas," then this protogyny, 
associated with other degradations of the corolla, etc., which 
only tend to increase it, has ended with anemophily. 

In the first case the androecium of protandrons flowers 
has come down from its previous highly differentiated state, 
so as to be homogamous with the stigmas. T^rom the other 
or protogynous condition, the gynoecium has not been brought 
back again so as to be homogamous with the anthers and 
pollen, but, on the contrary, it may have become even 
further differentiated, and so has now no fertiliser to depend 
upon except the wind. 



Degeneracy of the Andr(ECIum. — The number of stamens 
may decrease, as well as the quantity of pollen ; while the form 
of the anthers may change and the character of the pollen 
may alter ; and lastly, the position of the stamens may not 
be the same as in intercrossing flowers, — all these forms of 
degradation being so many adaptations or adjustments for 
self-fertilisation. They are well seen in Violets and the 

As examples, in Stellaria Holostea there are ten stamens, 
in S. media only three ; and in cleistogamous Violets they 
vary from five to three or two. In the latter, the anthers 
become spoon-shaped with a rounded connective and much 
reduced anther cells ; in the cleistogamous flowers of Oxalis 
Acetosella the pollen is almost deliquescent. Lastly, in all 
flowers especially adapted for self-fertilisation the anthers 
are in contact with the stigmas in consequence of their arrest 
in growth. 

It must be noted here that this degeneracy in the stamens 
in no way impairs their functional value. The fact is that 
a very small amount of pollen is really quite sufiicient for 
fertilising a considerable number of ovules. 

For convenience I call it degeneracy, but another view 
would be to regard it as the conservation of energy, instead of 


wasting it in the production of a gi'eat deal more pollen tlian 
is usually required. 

An interesting experiment of Mr Darwin's proves this. 
He placed a very small mass of pollen-grains on one side of 
the large stigma of IpomoBa purpurea, and a great mass of 
pollen over the whole surface of the stigmas of other flowers, 
and the result was that the flowers fertilised with little 
pollen yielded rather more capsules and seeds than did those 
fertilised with an excess.* That normally intercrossing 
flowers produce a great superfluity of pollen is well known. 
Thus Kolreuter found that sixty grains were necessary to 
fertilise all the ovules of a flower of Hibiscus, while he cal- 
culated that 4863 grains were produced by a single flower, or 
eighty-one times too many.f Mr. Darwin says, "In order 
to compensate the loss of pollen in so many ways, the anthers 
produce a far larger amount than is necessary for the fer- 
tilisation of the same flower; . . . and it is still more plainly 
shown by the astonishingly small quantity produced by 
cleistogene flowers, which lose none of their pollen, in com- 
parison with that produced by the open flowers borne by the 
same plants ; and yet this small quantity suffices for the 
fertilisation of all their numerous seeds." 

Mr. Darwin observed that when flowers were artificially 
self -fertilised for several successive generations, a degeneracy 
sometimes took place in the anthers and pollen ; and he seems 
to attribute this to what he called the " evil effects " of self- 
fertilisation ; but from the above-mentioned facts, which 
occur so abundantly in nature, I am inclined to regard it as 
an experimental verification and illustration of a universal 
principle in nature, namely the preservation of energy 
wherever possible, and that such cases^as appeared under his 

* Cross and Self Fertilisation of Plants, p. 25. 
t Ibid., pp. 376, 377. 


experiments were instances of tliis principle at work, as the 
flowers became habituated to self-fertilisation, and were then 
fully fertile. 

We have, then, in such cases an actual demonstration of 
the first step of the changes induced by self-fertilisation 
continiially enforced ; and thereby a witness to one cause of 
the origin of certain, and indeed, a very large number 
of species. It is the converse process to that of insect 
fertilisation, which itself I take to be the vera causa of the 
origin of intercrossing species. 

It is, perhaps, worthy of note that, while both the number 
of stamens and the quantity of pollen are thus often much 
reduced in some flowers the capsules of which produce many 
seeds, yet in others which set but one, as Fumaria, or at 
least but few seeds, the number of stamens may remain 
unaltered. This seems to me to be an additional proof that 
such flowers are degradations from forms originally adapted 
to intercrossing when much more pollen was requisite. 
Hence the present forms are retentions of former ancestral 
conditions. The following cases will illustrate this : — 
Scleranthus perennis £ind species of Medicago have ten stamens 
and one seed ; Daphne Laureola has eight stamens and one 
seed; Ghenopodium h.a,s five stamens and one seed; similarly 
is it the case with the large orders Composifce and Graminece. 

The phenomenon called " contabescence " by Gartner* 
would seem to have its rationale in this adaptation to self- 
fertilisation in some cases, and to diclinism in others, though 
there are other causes which may bring it about, when it is 
a purely pathological phenomenon. 

Mr. Darwin observes, " The anthers are affected at a 
very early period in the flower-bud, and remain in the same 
state (with one recorded exception) during the life of the 
* An. and PI. under Doni., ii., p. 165. 


plant. The affection cannot be cured by any change of 
treatment, and is propagated by layers, cuttings, etc., and 
perhaps even by seed. In contabescent plants the female 
organs are seldom aifected, or merely become precocious in 
their development. The cause of this affection is doubtful, 
and is different in different cases. . . . The contabescent 
plants of Dianthus and Verhascum found wild by Wiegmann 
grew on a dry and sterile bank." * 

" Cases of an opposite nature likewise occur — namely, 
plants with the female^ organs struck with sterility, whilst 
the male organs remain perfect." 

The constancy or prevalence of this condition of conta- 
bescence seems to be the first indication of diclinism, what- 
ever the cause ; and Silene inflata may be mentioned as 
frequently furnishing good examples of both kinds of 

Degeneracy of the Pollen. — As this is a feature of 
importance in the general degradation of flowers, a few 
words may be added in reference to it. It is of frequent 
occurrence in cultivated plants ; thus Potatoes are notorious 
for failing to produce fruit; and some varieties are much less 
liable to do so than others. Mr. C. F, White, F.L.S., tells 
me he regards this plant as furnishing the most conspicuous 
example of a form of degradation of pollen; the pollen 
grains of a normal character are very generally not to be 
found at all, but round, square, and polygonal forms abound. 
On the other hand, he gathered many flowers, in a large 
field in the Isle of Thanet, with scarcely a grain imperfect 
in shape or reduced in size. 

Mr. White has noticed, in his numerous researches 
among pollens, that degeneracy by dwarfing is mostly or 
very frequently induced by inclement weather. He mentions 
* A like cause produces petalody of stamens, see p. 299. 


the case of " Ononis, growing and flowering abundantly on 
the ' Sand-totts ' near Burnham, on the Bristol Channel, in 
which plant scarcely a grain of normal form was to be 
found ; many were absolutely united into grotesque groups 
and utterly deformed. . At the commencement of the cold 
weather of autumn, although the corolla may appear unin- 
jured, the pollen grains are often ' dirty,' unable, as it were, 
to throw oS. the residual tissue surrounding them, and are 
often irregularly reduced in size." 

This sensitiveness of pollen to barren soil, inclement 
weather, etc., at once throws light on a probable origin of 
diclinism, such as of gyno-dioeceous plants already mentioned; 
and simply confirms the idea that these differences in the 
sexual systems of plants must not be looked upon as so many 
beneficial arrangements, but simply inevitable results which 
must follow such circumstances as give rise to them, whether 
they may prove advantageous or not. The injurious effect 
of over-crossing, abundantly proved by florists, Mr. White 
recognizes in the character of the grains of Rhododendrons 
and Ericas, which exhibit a shrivelling up and occasionally 
a complete " dissolution " of one and the uppermost grain of 
the group of four. And this observer adds, that in more 
than one species of Erica and also of Vaccinium the injury, 
he thinks, has become chronic. 

If the " vegetative" system be too energetic the "repro- 
ductive " is sure to suffer, and one of the primary causes of 
the injury is the arrested state of the pollen, as Van Tieghem 
has described and figured it in Ranunculus Ficaria* A 
like result occurs in many cultivated plants, as Mr. Darwin 
has pointed out when describing the " contabescence of 
anthers." J 

* See above, p. 231, note. 

t An. and PI. under Dom., vol. ii., p. 165. 



DEGENERAcr IX THE Gyxcecium — If the theory be true 
that a tj^pical flower should contain two whorls of carpels, 
or, if spirally arranged, several cycles, then it is an obvious 
fact that these conditions are not the prevailing ones in 
nature. In a simple type, like JRanunculus, we find the pistil 
of many carpels, but with one ovule in each alone developed, 
except in monstrous conditions , if the ovules be numerous, 
then the carpels are reduced in number, as in the HellehorecB. 
This is a primary result of Compensation. And when 
carpels have become whorled —a condition I take to be 
primarily due to adaptations to insect agency, causing an 
arrest of axial growth by the enhancement of the corolla, 
etc., (see p. 6) — then degeneracy begins to play an important 
part, in that, firstly, (theoretically, be it observed) one of 
the two whorls of carpels goes altogether, sometimes the 
calycine (e.g. Fuchsia), at others the petaline (e.g. Cam- 

Secondly, the number of carpels diminishes, as in the 
Gamopetalce, where less than five prevail. The following 
table will show with tolerable accuracy the proportional 
number of carpels and ovules that prevail in the first three 
divisions of Dicotyledons. 

(1) Orders with many carpels or 

many ovules 

(2) Orders with 5 carpels and 

many ovules 

(3) Orders with 5 carpels and 

5-10 ovules 

(4) Orders with less than 5 carpels 

and less than 5 ovules 

(5) Orders with less than 5 carpels 

and many ovules 

Observations. — (1) The first-mentioned correlation has two 

Ord. p.c. 

Ord. p.c. 

Ord. p.c. 

12 or 19 

6 or 7 


12 or 19 

10 or 12 

7 or 12 

12 or 19 

14 or 17 

3 or 5 

14 or 21 

30 or 36 

23 or 40 

17 or 25 

22 or 27 

25 or 43 


conditions, either many carpels having one or few ovules 
in each, or a few carpels with many seeds, as in the 
RanunculacecB. This primitive condition rapidly vanishes in 
passing to Calycifloroe and Ga7nopetaIce. 

(2) Having reduced the nuriber of carpels to a definite 
quantity, five, i.e. one cycle of the prevailing -f type, this 
number remains tolerably persistent, but does not show a 
large percentage. 

(3) The combination of five carpels with a reduced number 
of ovules, i.e. one or two in each cell, or 5-10 ovules in all, 
is pretty uniform for the first two divisions, but almost 
disappears under Gamopetalce, the orders Sapotacece, Nolanece, 
and one or two Buhiacece, (e.g. EritJialis) representing this 

(4) and (5). Here we see a steady increase in the 
percentages in passing from Thalamijlorce to Gamojpetalce, in 
which the number of carpels is still further reduced ; but 
the number of ovules runs in two directions, being either 
numerous or few. 

Two questions arise at this point. If one result of insect 
agency is to bring about increased specialization in flowers 
(yet, in proportion as they become specialized, so, inversely, 
is the number and variety of insect visitors diminished), 
how is it that some (e.g. Foxglove and Orchids) produce 
an enormous number of seeds ; while others (e.g. Lahiafce, 
Compositce, etc.) produce few or only one in each flower ? 
The second question is whether a plant is better off for 
having so many more seeds than another. Recognizing 
reproduction as the sole end of plant life, so that a plant 
should bear as many good seeds as possible, it is noticeable 
that the two largest orders, Compositce and Graminece have 
never more than one seed to each flower. Again, comparing 
Lahiatce with Scrophularinece, according to the Genera 


Plantamm of Bentham and Hooker, while the former has 
2600 species, the latter has only 1900. Lastly, comparing 
two orders with regular flowers and two carpels, Boraginece 
has 1200 species, and Solanece, 1250 ; while the former order 
never has more than four seeds to a flower, in the latter they 
are numerous. 

If it were possible, we should procure statistics as to the 
relative degrees of abundance in individuals of two kinds at 
any place where they thrive. Casual observations certainly 
have not led one to notice any such jprojportional ahundance 
of the many-seeded plants as theoretically ought to exist 
if all their seeds germinated and grew to maturity ; for I 
have calculated the number of apparently good seeds in a 
large plant of Foxglove, and found it was one and a half 
millions. If we take a typical case, that of Orchids, whose 
flowers are certainly of those most highly adapted to insect 
agency, it is now well known that the proportion of seedlings 
to seed is infinitesimally fimall. Mr. Fitzgerald speaks of 
a Dendrohium speciosum, which bore 40,000 flowers open at 
the same time ; but though the plant was growing in the 
open air and was exposed to the visits of insects, only one 
-flower produced a seed pod* Mr. H. 0. Forbes found the 
same thing to occur in the terrestrial orchids of Portugal, 
and the tropical ones of Borneo. f Exactly the same diffi- 
culties are met with in cultivating plants, and especially 
Orchids (with few exceptions), as Mr. Yeitch has testified. 

Now, when we examine the structure of the essential 
organs of Orchids microscopically, their degeneracy at once 
becomes apparent. First, with regard to the pollen. Instead 
of its being in well-formed distinct grains, each with its 

* Referred to by Mr. Veitch, Beport on Orchid Conference, Journ. 
Roy. Hort. Soc. Bot., vol. vii., p. 47. 

t Journ. Lin. Soc. Bat, vol. xxi., p. 538. 


extine and intine, their development is arrested and, while 
still in contact, a common extine clothes the whole of each 
massula. Moreover, it is only after the pollen mass has 
been placed upon the stigma that the development is con- 
tinued.* With regard to the pistil the first sign of degeneracy 
is seen in the parietal placentation which prevails, and more 
especially in the rudimentary character of the ovules, every 
part of which is degraded. Even after fertilisation the 
enibrjo cannot grow to maturity, but remains in the arrested 
pro-embryonic condition. Having no albumen or nucellus- 
tissue wherewith to nourish the embryo, the suspensor does 
its best by elongating and escaping from the micropyle, and 
then, fastening itself like a parasite upon the placentas, ex- 
tracts nourishment therefrom — the result being that myriads 
of seeds never succeed (at least in cultivation) in developing 
even the pro-embryo ; and one can only infer that such is 
the case in nature. f 

In the cultivation of other flowers analogous phenomena 
are met with. The more highly cultivated a florists' flower 
may be, the less good seed is procurable; w^hile the poorer 
ones — that is, from a florist's point of view — or " w^eedy " 
looking plants furnish plenty, and are highly prolific. 

The rationale of these facts, whether taken from nature 
or from cultivation, I believe to be fundamentally the same, 
viz. the adaptation to insect agency and the result of repeated 
intercrossing, which enhances the development and form of 
the perianth especially, and generally of the stamens as well. 
At least the kinds of energy which are concerned in the 
manufacture of these whorls are more especially forced into 
activity by the stimulus received from without. On the 
other hand, the pistil suffers proportionately in all its parts 

* Mr. B. T. Lovme, Orchid Conference, etc., I.e., p. 48. 

t M. Guignard has drawn similar conclusions. See above, p. 172. 


through compensation and atrophy, the ovules being appa- 
rently particularly sensitive. To meet this diflBculty nature 
seems, to speak metaphorically, to have tried two methods, 
either to make an immense number of seeds, so that at least 
a few might be perfect, or else to attempt no more than four 
or even one, so that at least they should be vigorous, and 
survive in the struggle for life during the critical periods of 
germination and seedling existence. To judge by results, 
this latter method turns out to be the best. 

The interpretation, then, I would offer of inconspicaous- 
ness and all kinds of degradations is the exact opposite to 
that of conspicuousness and great differentiations ; namely, 
that species ^\iih. minute flowers, rarely or never visited by 
insects, and habitually self-fertilised, have primarily arisen 
through the neglect of insects, and have in consequence 
assumed their present floral structures. The external 
stimulus or irritations derived from the weights, pressures, 
and punctures of insects being no longer applied, the 
secretion of honey has failed, the corolla ceasing to be 
subject to hypertrophy has atrophied. A like procedure has 
obtained with the stamens, while a large proportion of pollen 
has become effete, the anthers being partly contabescent, as 
it is called. What remains, though often altered in cha- 
racter, is amply sufficient to set an abundance of seed. 

With regard to the pistil, however, the reverse of this 
has in some respects taken place. The corolla and andrcecium 
no longer putting a check upon the rapid development of 
the gynceciuni, the latter has a strong tendency to gain the 
ascendancy ; so that the result is homogamy or protogyny, 
with an extraordinary fertility of all plants which have 
inconspicuous and regularly self-fertilising flowers. 

If the seed be not always in great quantity in one and 
the same capsule, an ample progeny is secured by the 


extremely rapid maturation of the fruits in succession ; as 
may be remarkably well seen in Chickweed. 

The general result is that all these " weed-like " plants, 
with which wind-fertilised herbs must be associated as equally 
independent of insects, of all flowering plants are by far the 
most widely dispersed, and are, in fact, cosmopolitan ; * and 
although they be small and annuals, are yet best capable of 
holding their own in the great struggle for life. 

Rudimentary Organs. — These are the ultimate result of 
atrophy and degeneracy in flowers. They are so well known 
as occurring in all parts of plants, vegetative and repro- 
ductive, that I need not describe them now. The reader 
will doubtless gather from all that has been said about 
hypertrophy and atrophy as causes of development and 
degeneration respectively, that they are just what one 
would expect to find. Indeed, every organ can be met with 
in every stage of degeneration till it has completely vanished ; 
and even when all visible trace is wanting, the vascular cord 
belonging to it may in some cases still be detected. Last of 
all, this vanishes as well. These diiferences, for instance, 
can be witnessed in the presence or absence of the " trace " 
of the fifth stamen of the Labiates. 

It is thought by some that a rudimentary organ may 
become a honey-secreting gland, as Robert Brown suggested 
for some Cruciferous plants. Glands mostly consist of epider- 
mal and sub-epidermal tissues only, and if they occupy the 
place of an organ, the latter has the vessels arrested before 
they reach into the gland, which therefore is still of the same 
nature. In the male flower of Lychnis dioica the disk sur- 
rounds the rudimentary pistil, which in no way contributes 

* In my essay referred to, I have given a long list of self -fertilising 
plants which have been discovered in widely distant localities over the 
northern and southern hemispheres. 


to it. On the other hand, a gland may have its own proper 
vascular system, as in Lamium album, in which case a 
circular horizontal ring of vascular cords is formed from the 
pistillary cords ; from this are given off a series of vertical 
cords, r'jnning up into the gland itself. 

There can be no a priori objection to the supposition that 
an organ, when degenerating and becoming rudimentary, 
may acquire a new form and function ; for such, indeed, is 
not infrequently the case. But what perhaps may be more 
usual, is that some other organ becomes more highly de- 
veloped through compensation. Thus, for example, the 
leaflets of the Pea, in becoming tendrils, lose all trace of a 
blade, retaining only their mid-ribs. These, however, now 
elongate and acquire sensitiveness, for the use of climbing. 
On the other hand, in compensation for the loss of a certain 
amount of leaf surface, the stipules are very broad and 
foliaceous. Again, in the ray florets of Centaurea the essen- 
tial organs have vanished altogether, but the corolla is 
greatly enlarged in comparison with those of the disk 

* For a discussion npon " rndimentary organs," and their bearing 
upon the theory of Evolution, I would refer the reader to my work on 
Evolution and Religion (the " Actonian " Prize Essay for 1872), chap, 
xiii., p. 197. 



Homology. — The theory of homology has long been main- 
tained, and has met with such an overwhelming mass of 
evidence in its favour, that it is now regarded as a well- 
established morphological doctrine. The belief that every 
individual member of a flower, whether sepal, petal, stamen, 
or carpel, may be interchangeable with a leaf, and that they 
are therefore all phyllomes or foliar appendages to the axis, 
scarcely requires proof. Secondly, any one organ may 
theoretically be substituted for any other, so that although a 
sufficient number of interchanges has not yet been met with 
to make a complete series of permutations, yet they have 
gone far towards strengthening the probability that such 
might be possible.* 

I propose giving a very abbreviated series to illustrate, 
first, progressive changes from leaves through bracts to 

* The metamorphosis, with the exception of the substitution of 
petals for other organs, is rarely more than tentative ; for it is, as it 
were, a mere attempt to effect a change, so that wherever a " monstrous" 
organ bears ovules they are almost always rudimentary and quite 
incapable of being fertilised. I have said "rarely," for M. Brongniart 
succeeded in obtaining fertile seeds from artificial impregnation of 
ovuliferous stamens in Polemonium coeruleum (Bull. Soc. de Bot. Fr.y t» 
viii., p. 453). 



carpels; and, secondly, a retrogressive series from carpels to 
bracts, and thence to leaves ; finally deducing some important 

Progressive Changes in Bracts. — Bracts are in many 
cases very obviously modifications of leaves, being sometimes 
simply complete leaves reduced only in size, as in Epilohiuvi ; 
or a bract consists either of tbe blade alone, as in Buttei-cups, 

or else of the petiole only, 
but now expanded and 
blade-like in form, as may 
be well seen in Hellebores, 
where transitional states 
occur between the normal 
pedate leaf and true lan- 
ceolate bracts (Fig 61, 
a, h, c). 

When bracts are 
coloured otherwise than 
green, they then approach nearer to members of the repro- 
ductive or floral series rather than the vegetative, and in 
many cases are actually continuous in a spiral series with 
the sepals and petals, as in Cactus, Calycanthus, etc., and so 
assist in rendering the flower attractive. Several species of 
the genus Salvia, e.g. S. splendens, S. Bruantii, as well as of 
Bromeliacece, are remarkable for having brilliantly coloured 
bracts at the base of the flower. In some cases the bracts 
may be so arranged as to mimic a corolla, and indeed func- 
tionally replace it, as in species of Cornus (Fig. 62), Darivinia 
(Fig. 63), and the so-called Everlastings. 

The presence of bright colours in bracts, as also in sepals, 
to be described, I take to be due to the same influence as 
of the normal attractiveness in corollas ; viz., the visits 
of insects : the immediate cause being nourishment ; the 

Fig. 61.— Transitional forms, a, 6, from a leaf to a 
true bract, c, of Helleborus viridis. 



stimulus required to bring the extra flow to the bracts, etc., 
being presumably the irritation induced by insect visitors. 
The next progressive state is for bracts to assume a more 

Fig. 62. — Inflorescence of Cornus florida, 
with tour white petaloid bracts. 

Fig. 63.— Inflorescence of Darivinia, with 
coloured petaloid bracts. 

or less staminoid character. This is rare, but it has been 
noticed in Abies excelsa* A substitution of anthers for 
bracts has been seen, in Melianthus majo7\f concerning which 
Sig. Licopoli remarks that the flowers of chiefly the terminal 
racemes were imperfect, the summit of the floriferous axis 
bearing a tuft of perfect and imperfect anthers the petals 
and the two carpels of the flower having been atrophied or 

Fig. 64 represents an involucral bract of Nigella, bearing 
an anther on one side of it; while Fig. 65, a, is that of a 
glume of Lolium perenne with an anther. That bracts should 
ever assume a pistilloid character is, a priori, still more 
unlikely, as being further removed from the central organ of 
the flower. Dr. M. T. Masters has, however, described J a 

* Teratology, p. 192. f Bull. Soc. de Bat. Fr., Rev. hih., t. xiv., p. 253. 
X Journ. of Lin. Soc. Bat, vol. vii., p. 121. 



malformed Lolium perenne, in which the flowering glnmes 
had styles and stigmas (Fig. 65, a, h) ; the essential organs 
being absent, were replaced bj a tuft of minute scale-like 

Fis- 64.— Involucral bract of 
Nigella, with autber (after 

L'ig, 65.— Glumes of Lolium, with anther 
and stigmas (after Masters). 

organs, some of which were prolonged into stjliform pro- 
cesses, the sexual organs being otherwise suppressed. 

In a proliferous case of Delphinium elatum described 
and figured by Cramer,* the parts of the flowers were all 
metamorphosed into open rudimentary carpels. The axis 
was elongated and terminated above, in one case, by a 
similar abortive flower ; in another, by an umbel of such 
flowers, every part of which was more or less carpellary ; 
while all the bracts on the prolonged axis, even those out of 
the axils of w^hich the branches of the umbel sprang, w^ere 
similarly made of open carpels. 

Progressive Changes in the Calyx. — The sepals are 
usually homologous with the petiole of a leaf. This is obvi- 
ously the case with the Rose, where the rudiments of the 

* Bildungsahweichungen, etc., heft, i., taf. 10. The figure is repro- 
duced in Teratology, p. 126. 



compound blades are retained (see Fig. 24, p. 93). In 
Fedicularis the blades are present as a minute fringe on the 
edge. In Ranunculus, Poteiitilla, etc., the broad base of the 
petiole is the only part present, for in abnormal conditions 
the blade may be borne above (Fig. 6i5). Similarly, in a 
gamosepalous calyx the teeth as a rule seem to be all that 
remain to represent the blades ; for in TrifoUum repens, when 
virescent, true unifoliate blades are developed on elongated 
pedicels, all arising from the border of the calyx-tube (Fig. 
Q7)^ in which the teeth become pinnately nerved blades. 

Fig. Q&.—Rtnunculus with foliaceous 

Fig. 6Y. — Foliaceous calyx of TrifoUum 
repens, with stipulate leaflets (after 

The venation may in some cases assist in furnishing a clue 
as to the real nature of a part. Thus in Hellebore, as already 
seen (Fig. 61), the bracts are homologous with petioles, 
their venation being palmate, and not pinnate as in the 
divisions of the blades of the leaves. It is the same in the 
sepals, which are presumably therefore homologous with 
petioles as well. The sepals of Galtha resemble them in 
their venation, but in this plant the leaf is of a more 
primitive type, not being lobed, and has also a palmate 

A similar difference between the venation of the sepals 



and blades of the leaves is seen in Dlioterocarjpus and Mus- 
scenda (Fig. 68). Transitional states from a single to a 
doable flower of Saxifraga decipiens^ described and figured 
by M. C. Morren,* shows that the newly formed petals in 
the place of stamens, as also the normal petals of the flower, 
exactly correspond, both in shape and venation, with the 
cotyledons. Palmate venation thus simply represents a more 
primitive type; and, since flowers are constructed oufc of 
metamorphosed leaves — the vegetative being replaced by 
reproductive energies, — one naturally expects to find the 
calyx and corolla, which more nearly approach leaves in 
structure, to show arrested foliar con- 
ditions, as, e.g., are seen in palmate 
nervation and absence of blade or 
petiole, as the case may be. 

In Musscenda (Fig. 68) the teeth 
of the sepals are usually subulate and 
acuminate ; but in the one foliaceous 
and subpetaloid sepal it is drawn out 
into a long petiolar form, which then 
VVI /// \ \M /(/ expands into a palmately nerved 
nW \ n r // lamina. The fact that a "tooth" is 

1 1 \\ // ^^ ^^^^ ^^^® prolonged into a " petiole " 

-■' ^ '' seems to imply that the sepal arises 

at once from the receptacular tube, 
which, therefore, one would infer to 
be axial. A somewhat analogous pro- 
cedure is in the monstrous Trifolium, 
where the unifoliate blades, supported on long pedicels with 
stipular appendages as well, all arise from the border of- the 
so-called calyx-tube (Fig. 67). There the inference would 
be the same, only that the receptacular tube is free from the 

* Les Bull, de I'Acad. Roy. de Bruxelles, t. xvii., p. i., p. 415. 

Fig. 68.— Flower and leaf of 


pistil, and not adherent as in the case of Musscenda. In both 
instances it will presumably be purely axial in character. 

Progressive changes in the calyx are not uncommon by 
its assuming a petaloid character. This is normal in some 
genera of Banunculacece, in Fuchsia, BhodocJiiton, as well as in 
some members of the Incompleted, as in Mirahilis, Folygonuw, 
Daphne, etc. ]S"ormally coloured sepals are most frequent 
in polysepalous genera. Abnormal colorisation, with or 
without any metamorphosis of the organ, is 
most frequent in gamosepalous flowers, as in 
the cultivated " hose-in-hose " varieties of 
Primula, Miynidus and Azalea. The calyx 
may be petaloid either wholly or in part 
only. In Musscenda (Fig. 68), one sepal only 
is normally sub-petaloid. Calceolaria has 
occasionally one or more sepals petaloid. 
Similarly Linaria (Fig. 69) and other in- 
stances might be mentioned. These condi- 
tions, brought about by cultivation, clearly Fig. G^unaria. 
show the important part that high nourish- petaloid. 
ment plays as an external stimulus or factor in the produc- 
tion of colour. 

Staminoid sepals appear to be very rare. It is recorded 
by M. Gris that they have occurred in Philadelphus speciosus:* 

Pistiloid sepals are nearly equally as rare as staminoid. 
They have been observed by Mr. Laxton in double flowers 
of the Garden Pea (Fig. 70), in which there was a five or 
six-leaved calyx, some of the segments of which were of a 
carpellary nature, and bore imperfect ovules on their mar- 
gins, the extremities being drawn out into sub-stigmatiferous 


* Bull. Soc. de Bot. Fr., t. v., p. 330. 

t Gard. Chron. 1886, p. 897 ; and Teratology, p. 302. 



I have also found the sepals ovuliferous in a monstrous 
form of Violet, which was almost entirely virescent (Fig. 71). 

Progressive Changes of the Corolla. — For petals to 
become staminoid is far from uncommon. It is a normal con- 
dition in Atragene (Fig. 44, p. 141), which illustrates the 
transition, and in Water-lilies, where a gradual development 
of the anther cells is accompanied by a gradual reduction of 
the petal to a filament. As abnormal instances may be men- 
tioned, a case of Foxglove which I have elsewhere * described 
as having the corolla split up into strap-shaped antheriferous 
processes (Fig. 72), and a Columbine in which the spurs 

Fig. 70 Calyx of Garden F'ea. 

with carpellary lobes (after 

Fig. 71.— Ovuliferous 
sepal of Violet. 

Fig. 72.— Corolla of Fox 
glove, -with staminate 

became curiously coiled and bore pollen within the tissue of 
the coils (Fig. 73). 

Pistiloid petals are of rare occurrence. As an example is 
Begonia (Fig. 74, a), in which the apex of the petal was 
green and stigmatiform, the basal part being broad, coloured, 
and ovuliferous. Fig. 74, b, shows a petal, ovuliferous below, 
stigmatiferous at the summit, and antheriferous midway ; 
c is a rudimentary ovule. 

Progressive Changes in the Stamens. — The only change 
* Journ. Linn. Soc. Bot., vol. xv., p. 86, tab. 3. 



that stamens can undergo in tbis direction is to be more 
or less converted into pistillary structures. This is by no 
means uncommon. Either the filament alone, or the anther 
alone, or both together may be affected. The reader is 

Fig. 13.—Aquilegia, with polleniferous spurs 
(alter W. G. Smith). 

Fig. 74.— Ovuliferous petals, etc. 
of Begonia (after Masters). 

referred to Dr. Masters's Teratology for a description, -with 
figures of several kinds.* It is more usual for the filament 
to become enlarged into the ovarian part bearing rudimentary 
ovules ; but when the anther is involved, it may be partially 
or wholly transformed. In these cases the connective is 
usually prolonged into a stigmatiferous process. | As an 
example often described is that of the Houseleek, in which 
the margins of the anther cells become ovuliferous in various 
degrees ; as in Fig. 75, where ovules are borne by the pos- 
terior sides only, instead of pollen. In other cases the 
filament bears rudimentary ovules as well. Dr. Masters 
points out that " where there is a combination of the 

* Page 303. 

t In Aristolochia this change seems to be permanent and functional. 
See above, p. 83. 



attributes of the stamen and of the pistil in the same organ, 
the pollen is formed in the upper or inner surface of the leaf 
organ, while the ovules arise from the opposite surface from 
the free edge." Begonia is a genus which is peculiarly liable 
to produce malformations in the stamens (Fig. 76).* Rosa 

Fig. 75.— Ovuliferous anthers of 
Sempervivum (afier Masters). 

Fig. 76.— Stigraatiferous and ovuliferous 
stamens of Begonia. 

arvensis t affords a case in which the ovules were borne by 
the anthers, and then they themselves produced pollen. In 
these cases, where the anthers are ovuliferous, the connective 
is often more or less stigmatiferous, as in Begonia (Fig. 76), 
which shows various degrees of metamorphosis in this way ; 
but the anthers may sometimes be stigmatiferous, as in 
Poppies, X or styliform as well, as in Bamboos. § 

The complete substitution of carpels for stamens occurs 
in many plants, as in Malus apetala,\\ Tulips, etc., and is 
extremely common in Wallflowers,'|[ while it is by no means 
an uncommon occurrence to find male plants of normally 
dioecious or monoecious character bearing female organs, 
though perhaps in these cases it is often an addition, rather 
than a substitution of one organ for another. 

♦ See Journ. of Lin. Soc. xi. 472; Bot. Zeit. (1870), vol. xxviii., 
p. 150, tab. ii. 

t Journ. of Bot., 1867, p. 318, tab. 72. J Teratology, p. 304. 

§ Col. Munro, Trans. Lin. Soc, vol. xxvi., p. 7. 

(I Poiteau et Turpin, Arhr. Fruit., t. xxxvii., referred to by Moquin- 
Tandon, Teratologic, p. 220. 

% Called " Rogues " by the market -gardeners, as the corolla is want- 
ing or green. See Ann. des Sci. Nat., 5 ser., xiii., p. 315, pi. 1. 



The Pistil. — Commencing with tlie pistil, there may be 
changes in the ovary, ovules, style, and stigmas, separately 
or collectively. Instead of one or more 
ovules, a pistil may be formed within 
an ovary, as sometimes occnrs in Wall- 
flowers, Grapes, Oranges, etc.* A sin- 
gular instance is described by Dr. 
Masters t of a Carnation, " the placenta 
of which bore not only ovules but also 
carpels, the latter originating in a per- 
verted development of the former ; so 
that many intermediate stages could be 
traced between the ordinary ovule and 
the ovary. Some of these carpels, thus 
derived from ovules, themselves bore 
secondary ovules on a marginal pla- 
centa" (Fig. 77, a, carpel and section), the secundine, how- 
ever, being the only part developed (6). 

Stamens within an apparent ovary have occurred in 

» Teratology, p. 182. 

t i.e., p. 267. Perhaps the supposed " ovule within an ovule " may 
have been the nucellus only, more or less free from the secundine. 

Fig. 77. —Carpels and ovules 
on placenta ot Carnatioa 
(after Masters). 



Bceckia diosmcefoUa ; * but as tliej grew on the interior of 
the wall and not on an axile placenta, as is the normal con- 
dition in the Myrtacece, I expect that it was due to the 
staminal vascular cords branching off and coming out of the 
tissue within instead of at tlie summit of the hollow recepta- 
cular tube, the carpels being more or less arrested. A not 
uncommon instance is to find the pistils of Willows with 
open ovaries and bearing one or more anthers on the margins 
(Fig. 78, a). I have met with a similar occurrence in 
Banunculus auricomus (Fig. 78, b). Pistils of other flowers 

Fig. TS.— Stameniferous carpels of 'Willow 
(a) and Banunculus auricomus (Jb). 

Fig. 79.— a, Petaliferous placentas of Car- 
damine pratensis; b, of Rhododendron. 

have been known to bear anthers in a similar way, as 
Chamoerops humilis, Prunus,f etc. 

Pollen within ovules has been met with occasionally, as 
in Passiflora and Bosa arvensis-X 

In some members of the Cruciferce, as Cardami7ie pratensis 
(Fig. 79, a), round pods are formed instead of the usually 

* Teratology, p. 184. Possibly the ovary was entirely absent, and the 
stamens would then be growing on the interior of a closed receptacular 
tube, just as carpels grow upon the i aside of the hip of a rose. 

t See Weber, Verhandlung des Nat. Hist. Vereines der Preuss Bheiu' 
und Westph., 1860, p. 381. 

X Teratology f p. 185. 


long siliquas. These are full of petals, and if carefully 
examined appear to be whorled, with traces of stamens and 
pistil within them ; so that they represent flower-buds, but 
of which petals form the greater part ; similarly, Rhodo- 
dendrons and other flowers are known to bear imperfect 
flower-buds within the ovary (Fig. 79, h). 

Anthers occupying the place of stigmas appear to have 
occurred in Campanula,* Snowdrop, and double Tulips. 

The substitution of stamens for the entire pistil is of a 
less usual occurrence than the staminody of its parts : for 
cases, the reader may consult Masters's Teratology .f In a 
species of Orchis, probably 0. Morio, the ovaries were wanting 
altogether, a long pedicel taking their place, and within 
the reduced and regular perianth were two anthers on 
opposite sides (Fig. 23, a, p. 92), an apparent compensation 
in lieu of the pistil. 

The next and most frequent case of metamorphosis is 
that of conversion of carpels, and usually the stamens as well, 
into petals, or the so-called " doubling " of 
flowers. This is usually accompanied by a 
change from whorls to spirals with a multi- 
plication of the parts. Thus, in a double 
Wallflower, I have counted more than fifty 
petals spirally arranged. With regard to 
the petalody of the pistil, as Dr. Masters 
observes, " this is much less common than 
the corresponding change in the stamens. ^ft^.'.^d^J^^SS 
It generally affects the style and stigma jarpei of Foiyan- 
only, as happens normally in Petalostylis, 
Iris, etc." X Fig. 80 illustrates a metamorphosed carpel of 
Polyanthus, with a broad coloured appendage to the style. 

In some double flowers the carpels only are petaloid. 

* Teratology, p. 300. f Ihid., p. 299. % ^&^^-> P- 296. 



This has been observed in Anemone ^nemorosa, cultivated 
varieties of Ranunculus, Violet, and Gentiana Amarella. 

Retrogressive Metamorphoses of Stamens. — For the 
stamens to become petaloid, it is extremely common, as in 
double flowers, and such a change may represent what 
is normally the case in Water-lilies, Canna, and Atragene 
(Fig. 44, p. 141). Changes may apply to the anther lobes, 
connective, or filament, or to all together. Fuchsias often 
bear filaments with petaloid expansions of the apex, at the 
base of which are one or two anthers showing varying 
degrees of degeneration. This is a very similar condition to 
one in Petunia, described by Dr. Masters, in which the con- 
nective had developed into a green roundish blade bearing 
two anther cells at the base (Fig. 81).* In such cases, it 
seems to be the connective which has expanded outwards 
and become the blade of the petal or leaf. Similarly, in the 

Fig. 81.— Foliaceous connective of Petunia 

(after Masters). 

Fig. 82.— Petalody, or «' hose-in-hose " form, 
of counectives in a double Columbine. 

double Columbine petalody of the connective sometimes takes 
place (Fig. 82). j Commelina alha has also furnished a case 
of an anther lobe becoming petaloid. 

Causes of " Doubling." — There can be no doubt tliat 
petalody results from a weakened reproductive energy, espe- 
cially that of the andrcecium, which can become constitu- 
tional and may be hereditary and transmissible by crossing. 
* Teratology, p. 254. f Ibid., p. 293 


Cases seem clearly to show that a barren and dry soil, as 
well as a very dry atmosphere, are prominent causes for its 
appearance. Thus Mr. Darwin described a double Gentiana 
Amarella* growing " on a very hard, dry, bare, chalky 
bank." T. S. speaks f of a double Fotentilla as "grow- 
ing along a high wall, on a dry raised bank close to a 
beaten path, adjoining a gravelly field." Again, a writer in 
Gartenzeitung % alludes to the raising of double Stocks, and 
says that they should only have " just enough water for 
their preservation," and that " the starved state of the 
plants " causes doubling. He alludes to Camellias, also, as 
becoming double when grown in a dry soil. Kerria Jajponica 
becomes double in Europe, in consequence of its missing the 
wet season of Japan. It is well known that double flowers 
are more easily raised on the continent than in England, 
probably from a like cause, as our atmosphere is considerably 
more charged with moisture than a continental one. In 
raising double Stocks, it is customary to procure seed from 
the flowers on axillary shoots which have a weaker repro- 
ductive energy than those growing on the primary or central 
axis, the seeds being smaller and often misshapen. The 
above causes are, therefore, suggestive ; in that if a some- 
what elevated, dry, and poor soil, one devoid of phosphates, 
etc., be provided, the probability is that petalody will ensue. 
Having once shown a trace of the malady, florists know how 
to proceed in order to propagate and transmit the affection. 

There remains one other floral metamorphosis, and that 
is of petals into sepals. This condition approximates to 
virescence of the corolla, so that in many cases such a 
change could scarcely be called sepalody. But M. Godron 
has shown that when Eanunculus auricomus appears to be 

* Gard. Chron., 1843, p. 628 f Ibid., 1866, p. 973. 

+ Ibid., 1886, p. 197. 


apetaloas or to have a corolla consisting of a few petals only, 
it is due to the fact that the petals which are wanting are 
really present, but have become calycine. 

Origin of Homology. — Though we cannot penetrate into 
the arcana of life, nor trace the workings of its forces which 
bring about the development of any organ whatever, I 
think we can at least reach the physiological starting-point, 
so to say, of all these changes which I have briefly described. 
I have already mentioned that we may consider a vascular 
cord as the fundamental "floral unit," and as all cords are 
identical in character as long as they are within a pedicel, 
and, as far as one can observe, identical also in character 
even when they have penetrated the different organs, we at 
once see that there is a common source for each and all. 
Secondly, w^hen we trace these cords from the receptacle or 
axis into the floral members, we soon discover that any cord 
can supply two, three, or more totally different organs with 
their respective branches, as in the case of Campanula 
medium described above (p. 43). Indeed, starting, say, with 
five cords in a pedicel, they can supply any number of organs 
ad libitum, however diverse in character and however 
numerous they may be. Hence, although normally each 
whorl is stamped with its own individuality, it is easy to 
imagine, in accordance with the principles of evolution, that 
others may partake of it ; and so the characteristic features 
peculiar to one w^horl can transcend its limits, and influence 
others as well. 

Beyond some such interpretation as this, I do not think 
it is possible to go. 

In saying that a fibro- vascular cord can "give rise" to a 
sepal, or petal, or other organ, I need hardly remind the 
reader that I am only speaking metaphorically, in describing 
what one observes in studying the anatomy of flowers. 



Stamens. — The last changes to be described, which are 
common to all the members of a flower, are virescence, when 
thej retain their normal forms, but are simplj green ; and 
foliaceons conditions, when they assume more or less a truly 
leaf -like form. 

Dr. Masters has given descriptions t of several of each 
kind of floral members as well as of foliaceous bracts, to 
which I must refer the reader for details. There are certain 
particulars, however, to which I would especially draw 
attention as throwing light upon the ordinary structure of 
floral whorls, and especially that of ovules. 

Taking the Alpine Strawberry as an illustrative case, the 
petals, stamens, and carpels are often more or less foliaceous ; 
but the petals retain a palmate venation, though the three 
leaflets of the ternate leaf are pinnately nerved (Fig. 83, a, 
h). In the case of stamens the connective may be foliaceous, 
as in Fetunia (Fig. 81) ; J also in the Alpine Strawberry 
(Fig. 83, a) and in the " Green Rose " the anthers are often 
persistent on either edge of a leaf-like intermediate part 

* The abnormal assumption of a leaf-like cliaracter. 
t Teratology, p. 241, seqq. 
X Ihid., p. 254. 




(Fig. 83, c). A curious foliaceous modification is described 
by Miiller and figured by Masters,* in which the metamor- 

Fig. 83.— a, Foliaceous stamen, and b, petal of the Alpine Strawberry (after Le Maout 
and Decaisne) ; c, stamen of " Green Rose." 

phosed stamen had the appearance of two leaves united by 

their mid-ribs. It occurred in Jatrojpha Pohliana (Fig. 84). 
This will be alluded to again, as pecu- 
liarly significant. 

Phtllody of the Carpels and 
Ovules. — This is of much more frequent 
occurrence than of the stamens. The 
first condition of change is to leave the 
ovary open and to expose the ovules ; 
the style may still be stigmatiferous. 
The ovules then undergo phylloidal 
changes of different degrees, and much 
discussion has arisen as to whether the 

coats of ovules should be regarded as homologous with leaves, 

the nucellus being axial, or not, etc.f 

Since, however, anatomical observations clearly prove 

that both the primine and secundine issue out of tangential 

Pig. 84.— Foliaceous stamen of 
Jatropha Fohliana (after 

Teratology, p. 255. 

t L.c, p. 262. 


divisions of the epidermal layer of the nucellus, one can 
hardly consider these of themselves as homologous with a 
true phyllome. But when we find that each of the two sides 
of an anther cell can develop into a foliaceous structure, as 
in the case of the Jatrojoha alluded to above (Fig. 84), we 
seem to have discovered a power of converting what is 
originally and simply an epidermal layer into a truly folia- 
ceous structure. Moreover, this process is not infrequent 
in certain monstrous states of ovules, so that it would appear 
that any question of homology is, strictly speaking, out of 
court in these cases. When the whole of an appendicular 
organ becomes foliaceous, then, of course, a true case of 
homology may be recognized. 

Origik, Development, and Homologies of Ovules. — Tera- 
tology, here, I think, assists us greatly. With regard to the 
structure of an ovule, it first appears as a papilla upon the 
placenta, the cellular tissue of which, with its epidermal 
layer, constitutes the first stage. Such may, perhaps, be 
considered as the rudimentary condition of the funicle alone, 
as the true ovule is formed at the summit of it. One or 
more of the apical sub-epidermal cells gradually develop into 
the nucellus, while the sec undine is first formed by tan- 
gential division of the epidermis commencing at a certain 
place below the apex ; the primine, if present, subsequently 
following suit in the same manner.* It is a noticeable fact 
that while an ovule thus complete is elsewhere general in 
flowering plants, the Gymnosperms and most orders of the 
Gamopetalce form remarkable exceptions, as having only one 
coat to the ovule. In the former of these two groups it is 
doubtless due to a primitive condition being accompanied 
by other features showing affinities with cryptogams. In 

* See p<aper by Warming, De V Ovule, Ann. des Sci. Nat., v. (1877), 
p. 177. 


the latter, it is due to reversion by arrest, and is likewise 
accompanied by a simpler origin of the nucellus and embryo- 
sac, as Warming has shown. The suggestion I w:ould offer 
to account for this anomaly is that such arrest is due to 
compensation. The Gamopetalce, as a whole, are the most 
advanced of all flowers through adaptations to insect agency ; 
and as this invariably brings about an exalted condition of 
the corolla and stamens, the consequence is that the pistil 
has to suffer ; the first visible result being protandry, accom- 
panied by a temporal arrest in the development of the pistil. 

If this tendency to arrest be carried to the ovule, it may 
be affected too, and the result is that one, the last-formed 
coat, may be arrested altogether. Orchids, as shown above, 
illustrate this principle remarkably well, as their ovules, 
though possessing two coats, are as degenerate as in many 
parasitic plants (see above, pp. 172 and 281). 

Tracing the origin of an ovule, then, from its birth, it 
first appears as a papilla on the placenta of the carpel. A 
branch from the marginal fibro-vascular cord of the carpel 
enters it from below, and reaches at least as far as the 
chalaza, or base of the nucellus. It may go no further, as in 
rudimentary ovules of Orchis ; or be arrested in the form of 
cambium in the degraded state seen in the parasitic Thesium. 
On the other hand, as the ovule becomes a seed and the coats 
go to form the seed-skin, fibro-vascular branchings may occur 
all through the latter, being developed from the original cord. 
Such may be well seen in Mustard, Acorns, Beans, and the 

Although the coats of the ovule were originally formed 
by tangential division of the cells of the epidermis of the 
nucellus, when united to form a seed-skin, this has become 
thickened by a cellular growth between them, through which 
the cords then ramify. 



Pistils whicli have reverted to a more or less foliaceous 
character bear ovules which often become foliaceous as 
well ; and then a not uncommon procedarc is the develop- 
ment of a cup-like structure, probably composed of the two 
ovular coats, on an elongated stalk, with a rudimentary 
nucellus within, but more or less perfectly free from it ; or 
it may not exist at all. 

The late Professor Henslow described a monstrous con- 
dition of Mignonette with figures of ovules in this condition.* 
They were sometimes 
replaced by minute 
leaves (Fig. 85, c) ; or 
else in the place of 
each was a cup-like 
structure, elevated 
on a long stalk, with 
an egg-like nucellus 
within, but quite free 
from it. He likened 
it to the theca of a 
moss with its central 
columella. Comparing these two modifications, represented 
by Fig. 85, a and h with c, — or, again, those of Fig. 86, a 
and h, — the interpretation seems to be that the fibro-vascular 
cord passing up the funicle of the ovule becomes a petiole, 
and its prolongation constitutes the mid-rib. The secundine 
and primine with intermediate tissue become the blade, as 
seen in the foliaceous states of ovules, and constitute the 
" cup " when they assume that form. 

A similar process, I think, quite explains the origin of 
the foliaceous processes of the stamen of JatropJia, repre- 
sented by Fig. 84. The entire stamen is, of course, really 
* Trans. Camh. Phil. Soc, vol. v. 

Fig. 85. 

-Foliaceous and metamorphosed ovules of Migno- 
nette (after Prof. J. S. Henslow). 



homologous with one leaf alone ; but the membranes belong- 
ing to each anther, being of, at least, two layers of cells, 
have become foliaceous, just as the epidermis of the nucellus 
has done in the cases herein described; so that, in the 
Jatrojjlia, two leafy expansions were 
developed out of one. 
iV vLv l(Sl-^3 Other instances are known of ovules 

^ " ^ being represented by leaves, as Primula 

Sinensis^ Symphytum officinale* and 
Sisymbrium Alliaria (Fig. 86). 

Theoretically, it might be objected 
that a leaf (carpel) should give rise to 
a leaf (ovule or, at least, ovular coat) ; 
but foliaceous excrescences from a leaf- 
Fig. se.-Metamorphosed ovules surface are not at all uncommon, as, 
^;^L^rST^Tom7i for example, frequently occur in Cab- 
n(Ka, xvu., t. 20). bages,f where, in consequence of high 

nourishment inducing hypertrophy, any " rib " or " vein " 
may throw off a branch which can form a leafy expansion, 
which not at all infrequently becomes 
funnel-shaped, like the abortive ovules of 
the Mignonette. Similar funnel-shaped or 
tubular productions are found on corollas 
of semi-double flowers, as in Primulas, 
Cyclamens, Antirrhinum, etc., sometimes 
externally. Fig. 87 represents a like out- 
'?eiice'on the labeuim^of growth from the labcUum of Cattleya 
cattieyaMossiece. MossiecB ; and I have seen the posterior 

sepal of Vanda coerulea replaced by a pedicel with a cup at 
the apex exactly like the terminal process in Fig. 85, a. In 
all these cases I would regard such productions as due to 

* Teratology, p. 263. t I^id., p. 312, fig. 166, 


As another carious instance a remarkable form of the 
Sun- dew, Drosera rotuncUfolia, may be alluded to here, as 
throwing additional light upon the origin of ovules. It has 
been described and figured by Naudin,* and also by Plan- 
chon.| In this monstrosity the ovular coats were represented 
by " tentacles." These, as is well known, are not epidermal 
trichomes, but structures issuing from branches arising from 
the fibro-vascular cords of the leaf, and are therefore strictly 
homologous with the "funnels " on cabbage leaves.J 

The conclusion, therefore, which seems deducible from 
the foregoing observations is that an ovule is simply an 
appendage (not a bud) to the fibro-vascular cord of the 
margin of the carpel, and under monstrous conditions can 
grow into foliaceous excrescences to the carpellary leaf. It 
is not, therefore, axial in its character. Since all that is 
required to start from is a fibro-vascular cord, this may be 
furnished by any cord, even the mid-rib ; and such is the case 
in some monstrouc states of Primula, in which rudiments of 
ovules are found on the mid-ribs as well as on the margins of 
separate carpels. 

As the " funnels " on the mid-ribs and lateral veins of 
cabbage leaves are due to an abnormal condition of hyper- 
trophy, so ovules I consider as arising in a similar way, and 
take them to be due to the same influence, though of course 
it is normal in their case. The very presence of the large 
cords running up the margins of carpellary leaves, direct from 
the axis below, — being often, indeed, larger than the dorsal 
cord, — which then ramify, not only into each ovule, but often 
backwards within the carpellary walls till they reach and 

* Ann. des Sci. Nat., 2^ ser., vol. xiv., p. 14. 
t Ibid., 3® ser., voL ix., p. 86, tab. 5, 6. 

;J; The " pitchers " of Nepenthes, perhaps, originate in much the 
same way, from the original water-gland at the apex of the leaf. 


anastomose with the dorsal cord ; these, together with the 
greatly thickened cellular margins now constituting the 
placentas which supply the conducting tissue for the pollen- 
tubes, all show a form of hypertrophy in the edges of the 
carpellary leaves, a condition of things widely different from 
the usually thin and more or less impoverished margins of 
true leaves. 

If we may recognize a fibro- vascular cord as the " funda- 
mental unit," and as a basis, for the construction of any 
organ, and moreover as also containing within it poten- 
tially the power of evolving any number of similar organs by 
repeatedly branching ; then, when hypertrophy affects such a 
" unit," it may branch once, twice, or any number of times, 
when each branch passing off to the surface can lay the founda- 
tion of a repetition of the organ from which it takes its rise.* 
Attention has already been called to this origin of the 
numerous stamens of the Malvacece, and how certain forms 
of double flowers originate from the multi- 
plication of the petaline cords, each branch 
of which issues in a distinct petal, as in 

Similarly for carpels and ovules, the 
process of multiplication can be witnessed 
both normally and abnormally. On the 
Fig. 38.— Multifold carpels one hand, that of carpels into five groups 

■with ovuliferous margins . , , 

from a malformed Prim- occurs in the Hoilyhock througn the 
chorisis of the original carpellary cord ; on 
the other. Fig. 88 represents a multifold carpel of a Prim- 
rose, due, there is no doubt, to a like chorisis of the cords 
belonging to one individual carpel. 

Similarly for ovules, while two only are normally charac- 

* I must again remind the reader that I am here speaking meta- 
phorically ; as we do not know wherein this potentiality really lies, but 
can only describe what is actually visible. 


teristic of the Plum, and Orcliis has an innumerable quantity 
arising by repeated chorisis of the original placentary cords, 
so in monstrous Primroses, etc., as represented in Fig. 88, 
many additional ovular processes may be formed, not only on 
the margins, but even, as stated, on the mid-ribs. 

One other point may be here noticed, a jpropos to the 
following curious discovery by M. Baillon. I have regarded 
a vascular cord as a " unit," as being capable of giving rise 
to any appendage whatever ; and as long as it is in an axis 
as a "trace," the cords of all organs are absolutely indistin- 
guishable. Further, there is no difference between a cord 
which will enter an appendage and one which will form a, 
pedicel from a peduncle. In the latter case, several cords 
are usually required for the pedicel ; while one, the most 
external of the " horseshoe " group given off at one side of 
the peduncle {i.e. as seen in a transverse section), enters the 
bract. In Erodium cicutarium, which has three flowers to 
the umbel with very slender pedicels, one single cord is all 
that the peduncle contributes to supply each of the pedicels, 
and one very small cord for a bract. The cord for the pedicel 
increases by radial chorisis, and so passes from the form of a 
wedge to that of a fan, when the outermost parts increase 
till they meet, and so a circle is established. 

This shows that an "axis" and an "appendage" are 
fundamentally due to the same kind of " unit." 

The reader will now see that in the following case * 
the funicular cord, which is normally that of a foliar, i.e. 
appendicular organ, supplied an axial cord ; just as many 
leaves can give rise to buds which are often utilized for pro- 
pagative purposes. 

* Sur le DSveloppement et la Germination des Grains Bulhiformes des 
Amaryllidees, par. M. Baillon, Bull. Soc. de Fr., t. xxi., 30 (pub. en Revue 
des Cours Sci. Lyon, Aoiit 30, 1873). 


" Les bulbilles des Amaryllidees ne sont pas toujours des 
graines veritables modifiees seulemenfc quant a I'epaisseur et 
a la consistance de leurs diverses couches natarelles, notam- 
ment des plus exterieures. Temoin le Calostemma Gunning- 
hamii. Ici, par une singuliere transformation de I'ovule en 
bulbe, la chalaze, en s'epaississant, joue le role d'un veritable 
plateau, sur lequel se produisent une, puis plusieurs racines 
adventives. Les enveloppes ovulaires tienne alors lieu 
d'ecailles bulbaires tandis qu'il s'eleve dans le sac embryon- 
naire un veritable bourgeon parti de la chalaze comma 
support et s'echappant par son sommet de la cavite ovulaire 
pour se comporte ensuite comme une plante complete." 

An analagous case of bulbs arising from a foliar organ 
occurred with Scilla Sihirica. Some plants dug up in October, 
1887,* were found to have taken the form of the so-called 
" droppers " not uncommon in tulips. Their peculiarity 
resides in the fact that the tubular leaf-sheath bends and 
grows downwards, thereby carrying the axillary bulbil to a 
greater depth in the soil than usual. In February, 1888, on 
re-examining them. Dr. Masters discovered that from one to 
four bulbils had been developed at different heights within 
the tissue of the tubular sheath, being in connection with the 
cords of the latter by means of a transverse nexus of tracheids. 

I refer to these cases as being curious instances of axial 
structures proceeding from foliar — i.e., by the change of 
character of the fibro-vascular cords from being at first 
foliar and then axial. They support the theory of homology 
of leaf and axis, w^hich is, of course, otherwise quite efficiently 
substantiated by such plants as Xyloiohylla, Buscus and the 

* Gard. Chron. for Oct. 15, 1887, p. 475, fig. 98, " droppers ; " also, for 
March 3, 1888, p. 276, fig. 45, ditto, with bulbils. 



There are at least seven kinds of union : — (1) self-fertilisa- 
tion, or the fertilisation of a pistil by the pollen from the 
same flower (Autogamy) ; (2) crossing different flowers on 
the same plant ; (3) crossing flowers on different plants of 
the same stock ; (4) crossing flowers of different plants, but 
of different stocks ; all the preceding being of exactly the same 
form or variety of species ; (5) crossing varieties of the same 
species ; (6) crossing different species of the same genus ; 
(7) crossing different genera of the same order. 

When a knowledge of the floral sexes was first acquired, 
the idea maintained was that hermaphrodite flowers were 
specially adapted for self-fertilisation ; but it was, I believe. 
Dean Herbert who first observed the importance of crossing, 
in his work on the Amaryllidacece (1836). He says, " I am 
inclined to think that I have derived advantage from im- 
pregnating the flowers from which I wished to obtain seeds 
with pollen from another individual of the same variety, or, 
at least, from another flower rather than its own, and espe- 
cially from an individual grown in a different soil or aspect." 

Mr. Darwin's work. On the Cross and Self Fertilisation 
of Plants (1876), placed on a scientific basis, -by means of 
experimental veriBcations, the exact values of such crossings. 
His conclusions, however, require considerable modifications. 


They are true for at least the first few years ; as in but four 
or five cases only did he exceed the third generation ; 
and when he prolonged them to seven generations, as in 
Mimulus luteus, and ten in Ipomcea 'purpurea, his results 
began to assume a very different complexion. 

The inference deducible from his experiments is that 
careful and artificial crossing generally introduces a remark- 
able stimulus for a time ; but the effects are not permanent. 
On the other hand, a perseverance in self-fertilisation pro- 
duces results which are much more stable ; so that, finally, 
self-fertilised plants {i.e the successive off-spring of this 
process) outstrip their competitors. Florists also find that by 
continued crossing the flowers of a species they soon reach 
the end of their tether, and no further progress is obtainable. 

Secondly, Mr. Darwin failed to realize the fact that self- 
fertilisation predominates in nature with the vast majority of 
hermaphrodite plants, whether they be adapted to insects or 
are inconspicuous and adapted for autogamy. 

Thirdly, he misinterpreted the meaning of degeneracy, 
which often accompanies self-fertilisation ; thinking that it 
involved constitutional injuriousness, of which there is no 
trace whatever in nature. 

Lastly,* he, and other writers who have followed him, 
wrongly inferred that adaptations to insect agency implied 
a converse "purpose," viz. to avoid self -fertilisation, instead 
of regarding them as the inevitahle results of the stimulus of 
intercrossing and of the visits of insects. The danger of 
this a ijriori, deductive, or teleological reasoning, without any 
attempt at verification, lies in the fact that it is untrust- 

* I mnst refer the reader to my paper on The Self-fertilisation of 
Plants, in which I have dealt with these points. It was written in 1877, 
but I have met with nothing since to invalidate the above conclusions, 
but, on the contrary, very much in support of them. 


worthy. It may or may not be true ; but it is of no value 
unless thoroughly tested by experiment aud verified. Thus, 
for example, Mr. Darwin, in speaking of the movements of 
the stigmatic lobes of Mimulus, says " Mr. Kitchener has 
ingeniously explained the use of these movements, namely, 
to prevent the self-fertilisation of the flower." He, however, 
experimented with this plant, and then discovered that " if 
insects are excluded the flowers fertilise themselves perfectly, 
and produce plenty of seed." * 

Again, it has been argued that we are justified in assum- 
ing that the remarkable adaptations to insects, which are so 
obvious in many flowers, viust he of some use to the plant, 
even though we may not be able to discover it. This state- 
ment, however, is just as much an a priori and deductive 
assumption as the preceding, and is quite valueless until 
verified ; and it is only by means of such experiments as Mr. 
Darwin laboriously carried out, that the real value of inter- 
crossing and self-fertilisation or other kind of union can be 
ascertained. Thus, e.g., the Garden Pea is undoubtedly 
adapted to insects, like other irregular flowers ; but experi- 
ments proved that " a cross between two individuals of the 
same variety does not do the least good to the offspring, 
either in height or fertility." f 

* Ci-oss and Self Fertilisation, p. 64. As another instance of an a 
priori deduction, Sachs says of Epipactis latifolia, " The flower left to 
itself does not get fertilised, for the pollen-masses do not spontaneously 
fall out of the anther; and even if they did, would not come on to the 
stigmatic surface " (Veg. Phys., p. 796). Mr. A. D. Webster, however, has 
observed that E. latifolia is very imperfectly fertilised, for, although 
visited by insects, cross-fertilisation seldom takes place ; that " self- 
fertilisation by the pollen falling spontaneously on the stigma is not 
nncommon, as the pollen-masses . . . become friable, and before the plant 
withers, either spontaneously or by the action of the wind fall on the 
stigma" {Bot. Gaz., xii., p. 104). 

t i.e., p. 264. 


Moreover, however greatly we may feel impressed with 
the truly wonderful adaptations of flowers, a careful and 
critical study of them reveals many features which seem 
to counterbalance, to some degree at least, the " good " we 
may in the first instance be inclined to assume as self- 
evident. Indeed, the disadvantages accruing from great 
differentiations in adaptation to insect agency are really too 
important not to have been frequently noticed. Such are, 
" hercogamy," or the mechanical obstruction to self-fertilisa- 
tion, as in Orchids ; the physiological barrier, as in Linum 
perenne ; the absence of insects required to fertilise a flower, 
as is the case with Convolvulus sepiiim in England, which 
rarely sets seed, as Sphinx Convolvuli is a rare insect ; the 
frequent absence of bees, etc., in inclement weather, when 
Clover sets but little seed, to the great loss of the farmer ; 
when certain flowers are neglected for greater attractions, 
as may be often seen when bees keep persistently to one 
species of plant and pass over others ; the frequency with 
which bees perforate tubular flowers without pollinating 
them at all. Again, MuUer points out* that Avhile honey- 
seeking insects may legitimately cross heterostyled plants, 
pollen-seeking insects have no need to thrust their heads or 
proboscides down to the stigma of the short-styled forms ; 
hence such tend to bring about illegitimate unions of the 
long-styled forms only. This, he thinks, may be a cause of 
the greater fertility of that kind of union. | Lastly, the 
more highly differentiated a flower is, the less is its number 
of insect visitors and the rarer may it become in nature. 
Thus orders of plants Avith easy access to the honey are 
some of the most abundant, as EanunculacecB, Compositc8,X 

* Fertilisation, etc., p. 387. f See above, p. 206. 

X The enormous number of species and '•vide diffusion of the Com- 
positcB are proofs of the advantages accruing to it from tie peculiar 


and Umhelliferce ; as well as are tliose dependent upon tlie 
wind, which never fails, such as Willows, Cyperacece, and 
Grasses. On the other hand, all regularly self-fertilising 
plants are abundant, and, together with certain wind- 
fertilising plants, are cosmopolitan. 

Although the idea that self -fertilisation is injurious is 
certainly not held now by botanists in so absolute a form as 
Mr. Darwin often stated it, yet it will not be amiss to point 
out the want of agreement between his conclusions and his 
own experiments. 

In a chapter on " General Results," * he commences by 
saying : " The first and most important of the conclusions 
which may be drawn from the observations given in this 
volume, is that cross-fertilisation is generally beneficial, and 
self -fertilisation injurious. This is shown by the difference 
in height, weight, constitutional vigour, and fertility of the 
offspring from crossed and self-fertilised flowers, and in 

structure of the flowers ; first, in being adapted to a great variety of 
insects. Thus, on ten species of plants, Miiller detected 546 species of 
insects, in the following proportions, Lepidoptera, 15 p.c. ; Apidse, 41 
p.c. ; Diptera, 27 p.c. ; other short-tongued insects, 17 p.c. Bees, 
therefore, are the chief visitors. This is almost invariably the rule : the 
only species mentioned by Miiller in his table in which short-lipped 
insects surpass in number the Apidee is Chrysanthemum leucanthemum, 
which has a corolla tube, 3 mm. in length, in which the honey rises 
up into the widening throat and is easily accessible. The number of 
Lepidoptera is in the proportion of 6*9 p.c. ; Apidae, 16*6 p.c. ; Diptera, 
38*9 p.c. ; others, 37'5 p.c. In Achillea Millefolium, vfith. a corolla tube 
of 3 mm., Lepidoptera are 6"9 p.c; Apidae, 34*5 p.c; Diptera, 24*1 
p.c, and others, 34*5 p.c. Lastly, in Centaurea Jacea, with a tube of 7 to 
10 mm., the Lepidoptera rise to 27 p.c. ; Apidse, 58*7 p.c. ; while Diptera 
sink to 12 '5 p.c, and other short-lipped insects are only 2 p.c. 

The Compositce thus well illustrate the fact that tubes are propor- 
tionate in length to the more specialized insects, a universal feature 
seen in all other orders as well. 

* Cross and Self Fertilisation of PlantSf p. 436. 


the number of seeds produced by the parent plants. With 
respect to the second of these two propositions, namely, that 
self-fertilisation is generally injnrions, we have abundant 
evidence. The structure of the flowers in such plants as 
Lobelia ramosa, Digitalis purpurea, etc., (1) renders the aid 
of insects almost indispensable for their fertilisation ; and 
bearing in mind the prepotency of pollen from a distinct 
individual over that from the same individual, such plants 
will almost certainly have been crossed during many or all 
previous generations. So it must be, owing merely to the 
prepotency of foreign pollen, with cabbages and various 
other plants, the varieties of which almost invariably in- 
tercross when grown together. The same inference may 
be drawn still more surely with respect to those plants, such 
as Reseda (2), and EscliscJioltzia (3), which are sterile with 
their own pollen, but fertile with that from any other 
individual. These several plants must therefore have been 
crossed during a long series of previous generations, and the 
artificial crosses in my experiments cannot have increased 
the vigour of the offspring beyond that of their progenitors. 
Therefore the difference between the self-fertilised and 
crossed plants raised by me cannot be attributed to the 
superiority of the crossed, but to the inferiority of the self- 
fertilised seedlings, due to the injurious effects of self- 

Mr. Darwin then proceeds to discuss the first proposition, 
*' that cross-fertilisation is generally beneficial," so that we 
may conclude that the preceding quotation represents the 
author's reasoning and conclusions on the idea of there 
being some " injuriousness " in self -fertilisation. 

In the first place, it may be observed that the reason 
why Mr. Darwin's crossings yielded at first more marked 
results in height, fertility, etc., is because plants are never 


SO carefully crossed in nature, nor self -fertilisation so carefully 
prevented, as was the case in his experiments. The probability 
is that the two processes are much more mixed in nature in 
the case of most plants. Therefore, by his experiments the 
more unalloyed influence of crossing brought about a much 
more enhanced stimulus than ever occurs in the wild state. 
Moreover, the prepotency of foreign pollen, upon which he 
lays stress, is a purely relative phenomenon ; for whenever 
self-fertilisation yields more seeds than intercrossing, as is 
often the case, it is a just inference that the pollen " of the 
same flower " is then prepotent, in its turn. Indeed, Mr. 
Darwin actually found that in some cases intercrossing did 
no " good " at all, as in the case of the Garden Pea mentioned 
above, and in Canna Warscewiczi, etc. 

I will now add some observations upon certain points 
I have numbered in this paragraph. 

(1) That Lobelia ramosa and Digitalis purpurea, and many 
others given in a "List of Plants Sterile without Insect-aid," * 
cannot readily fertilise themselves unless the flow^er be 
disturbed in some way, is, per se, no proof that self-fertilisa- 
tion is injurious ; for the flowers of many of such plants are 
fully self-fertile when artificially assisted. Thus, Mr. Darwin 
says that although Lupinus luteus and L. pilosus seed freely 
when insects are excluded ; yet Mr. Swale, of Christchurch 
in !N^ew Zealand, found Lupins only formed pods of seed 
when the stamens were artificially released, as they are not 
there visited at all by bees.f The interpretation of this 
fact, so w^ell known that the term " hercogamous " X has 
been invented for it, I take to be an immediate result of 

* Cross and Self Fertilisation, etc., p. 357. 
t L.c, p. 150, note. 

J If I remember rightly, by Errera ; see Bull, de la Soc. Bot. de Belg.y 
xvii. (1887). The term means a " fenced-off union." 


the action of insects. I have given reasons for believing, and 
the reader can readily suggest other instances, that structural 
peculiarities have grown in response to pressures and thrusts 
made upon the floral organs by the insects themselves ; 
and that such have sometimes produced protuberances or 
obstructions in the way of the emission of the pollen upoa 
the stigma of the same flower, is no more than might be 
anticipated to be extremely probable. Thus one of the 
most remarkable is the rostellum of Orchids, believed to be 
a modified stigma now converted to a new use. In nearly 
all Orchids this blocks up the way of access to the stigmatic 
chamber, while the pollen masses recline on the roof over 
it, so to say; but when Orchids become self-fertilising or 
even cleistogamous as well, this is often brought about by the 
degradation of the rostellum ; so that the pollen masses can 
then easily slide over the summit of the stigmatic chamber 
and fall into it at once. When they do so they are fully 
self-fertile, as Mr. Henry 0. Forbes has shown.* 

Some few plants are quite barren with their own pollen, 
even when artificially placed upon the stigma ; though 
Lobelia and Digitalis do not belong to the group. These, 
as shown elsewhere, can and often do become fully fertile 
at other places and seasons, and are thereby benefited by 
acquiring the possibility of setting seed by self-fertilisation, 
as otherwise they might set none at all. 

There are, then, three kinds of barriers to self-fertilisation : 
one mechanical, as in Orchids ; a second, that of time, when 
e.g. a flower is so strongly protandrous that the pollen is all 
shed before the stigmas are mature ; and, thirdly, a physiolo- 
gical one, when the pollen is actually impotent on the 
stigma of the same flower, even though it be homogamous. 

* On the Contrivances for insuring Self-fertilisation in some Tropical 
Orchids, Journ. Linn. Soc, xxi., p. 538. 


In no case is it logical to say tliat sucli arrangements are to 
prevent self-fertilisation. We may well ask why are a 
comparatively few plants thus provided for, and yet the 
vast majority are not. If, however, we regard them as 
results of differentiation brought about by the stimulus of 
insect agency — so that in certain places hypertrophy has set 
in and rendered the flower hercogamous, in others the 
androecium is so stimulated and its develoiDment so hurried 
on that the flower becomes protandrous, or its pollen so 
highly differentiated as to become like that of a distinct 
species, — we have a reasonable interpretation for these 
phenomena. Moreover, not one of them is absolute or stable. 
Thas a hercogamous Orchid can become self-fertilising;* 

* Since the above was in type, Mr. H. N, Ridley has read a paper, 
at a meeting of the Linnean Society (Feb. 16, 1888), on "The Self- 
fertilisation of Orchids," in which he arrives at the same conclusions 
as Mr. H. O. Forbes (see above, p. 253, note), finding that the process 
is effected in several ways, especially, perhaps, by the degeneration of 
the rostellum. Moreover, the Orchids which he discovered to be capable 
of fertilising themselves are not only the most numerous in individuals, 
but are also the most widely dispersed of the genera to which they 
respectively belong. He also corroborates Mr. Forbes's observations, that 
Orchids set but a small percentage of their fruit, although fully exposed 
to the visits of insects. 

Mr. H. Veitch has also contributed a valuable paper on the 
"Hybridisation of Orchids," in which he appears to corroborate M. 
Guignard's observations in every particular (see above, Chap. XVIII.). 

The reader will take note of the significance of the fact that when 
Mr. Darwin published his work on " The Fertilisation of Orchids," it 
was thought that no flowers could equal them in their remarkable 
adaptations for securing the benefits of intercrossing by insect agency, 
and in their methods of " preventing self-fertilisation." Yet, of all 
flowering plants, evidence now tends to show that they set the least 
amount of seed, even when fully exposed to insects ; while the order 
has furnished materials for two important papers on the many forms 
and ways by which self-fertilisation is secured in different genera. 


the strongly protandrous Carnation can be made to be 
liigbly self-fertile, as Mr. Darwin showed; and Linum perenne 
can have its pollen so modified as to set seed abundantly in 
the same flower, as occurred w^ith Mr. Meehan in Philadephia, 
though it was physiologically impotent in England. 

It is, in fact, so to say a mere accident that mechanical 
and physiological barriers exist at all ; and it is only by 
experiment that one can discover whether a flower so 
conditioned may not be really capable of self-fertilisation all 
the time. Indeed, Mr. Darwin's experiments have abundantly 
shown that self-fertilising properties are quickly reacquired, 
whenever the process is persevered with. For example, 
E sells choUzia Ccdifornica was "absolutely self -sterile " in 
Brazil. Mr. Darwin, however, by self-fertilising it in Eng- 
land, raised the fertility in two generations to nearly 87 p.c. 

When he asserts that his artificial crossings could not have 
increased the vigour of the offspring, and therefore all differ- 
ences must be attributed to the inferiority of the self-fer- 
tilised, this argument would apply to a certain number of his 
experiments in different degrees, viz., with plants normally 
self-sterile ; but he ignores the fact that, as soon as he tried 
to raise a stock of self-fertilised plants, the latter steadily 
gained upon the offspring of the crossed, till they equalled 
or surpassed them, or else would have done so had the 
experiments been continued. 

Thus, with regard to Lohelia ramosa, the ratio of heights 
of the " intercrossed " to the " self-fertilised " offspring of 
first generation was 100 : 82 ; and the proportion of seeds as 
100 : 60. In the second year, those growing under what he 
had proved to be the most disadvantageous condition for 
self-fertilised seedlings, namely, being crowded, the ratio 
of the heights became as 100 : 88-3. The experiment, 
unfortunately, was not continued further. 


Comparing tliis plant witli Ti. fulgens, whicli is also quite 
sterile without aid, and, according to Gartner, is " quite 
sterile with pollen from the same plant, though this pollen 
is efficient on any other individual," * Mr. Darwin suc- 
ceeded in raising self-fertilised plants by keeping the pollen 
of a flower in paper till the stigmas were ready, as it is 
strongly protandrous. The heights of the offspring were as 
100 : 127, and Mr. Darwin adds, " the self -fertilised plants 
[in two out of four pots] were in every respect very much 
finer than the crossed plants." 

In the next generation he used pollen from a different 
flower on the same plant to represent self -fertilisation. In 
this case those " self-fertilised " were only 4 p.c. below the 
crossed, the ratio being as 100 : 96. The conclusion, then, is 
that self-fertilisation pure was the best ; intercrossing distinct 
plants, less so ; and crossing on the same plant, the least. 

DiantJms, like Lobelia fulgens, is strongly protandrous ; 
but in the third generation the proportional number of seeds 
per capsule was as 100 : 125. " This anomalous result is 
probably due to some of the fertilised plants having varied 
so as to mature their pollen and stigmas more nearly at 
the same time than is proper to the species " (p. 135). 
Exactly so. 

The conclusion I would draw is, therefore, not that self- 
fertilisation is per se in any way injurious, but that flowers 
which are normally sterile, by having become so highly 
differentiated through insect stimulation, do not now spon- 
taneously set seed ; and self- fertilisation is not so efficient 
as crossing. As soon, however, as the former process is 
persevered with, signs are not wanting of nature's showing 
even an eager response to it, till the results are often far 
superior to those normally obtained by intercrossing. 
* Cross a7id Self Fertilisation, etc., p. 179. 


If flowers, unlike the preceding, are normally very self- 
fertile, as IpomcBa and Mimulus proved to be, then it appears 
that intercrossing supplies a remarkable stimulus, and the 
intercrossed beat the self-fertilised for a time. Sooner or 
later, however, the effect of the stimulus gradually dis- 
appears, and self -fertilisation, reasserts itself. Thus with 
Ipojncea purpurea Mr. Darwin raised crossed and self-ferti- 
lised plants for ten generations ; and the heights of the 
latter were 24, 21, 32, 14, 25, 28, 19, 15, and 21 p.c.,* respec- 
tively, less than the crossed. Grouping these into threes, 
the ratios become 100 : 74-3 ; 100 : 77-6 ; 100 : 81-6. That is to 
say, the intercrossed were steadily declining ; for if the self- 
fertilised be regarded as 100, then the ratios of these to 
the crossed appear as follows : 100 : 134 ; 100 : 129 ; 100 : 121 
Similarly with regard to fertility, the ratio of that of the 
intercrossed plants to the self-fertilised was for the first and 
second generations as 100 : 93 ; for the third and fourth, as 
100 : 94 ; for the fifth, as 100 : 106 ; and the eighth, as 
100 : 113. Hence the self-fertilised were superior. 

Mimulus luteus gave analogous results. The crossed 
plants {i.e. offspring of crossings) surpassed the self-ferti- 
lised until the fourth generation, when several plants of the 
latter assumed a taller character, with whiter blossoms. 
This self-fertilising form "increased in the later self -ferti- 
lised generations, owing to its great self- fertility, to the com- 
plete exclusion of the original kinds." f " It transmitted its 
character faithfully, and as the self-fertilised plants consisted 
exclusively of this variety, it was manifest that they would 
always exceed in height the crossed plants." X 

* These numbers correspond to the first nine years. The tenth gives 
46; but Mr. Darwin thinks this number to have been accidental (p. 41). 
t Cross and Self Fertilisation, p. 67. 
X Ibid., p. 70. 


(2) Witli regard to Reseda and Eschscholtzia, his observa- 
tions are also somewhat misleading. Mr. Darwin experi- 
mented with B. lutea and U, odorata. They are both very 
capricious. Of B,. lutea some individuals were absolutely 
self-sterile, whether left to themselves or artificially polli- 
nated, while a few produced self- fertilised capsules. Simi- 
larly with jK. odorata, when protected by a net some plants 
were loaded with self-fertilised capsules, others produced a 
few, and others, again, not a single one. Miiller,* however, 
found that " plants which are kept protected from insects, 
yielded capsules filled with good seed." The inference from 
this variability in the fertility of different individuals in the 
same year, is that it is an accidental peculiarity of some to 
be more or less self-fertile than others ; and that it was due 
to varying degrees of nutrition affecting the essential 
organs. We know now that plants frequently vary in their 
degi^ees of fertility, both at different seasons of the year,t 
and in different years or localities, according to climate, con- 
ditions of soil, etc. In any case, the self-sterility of these 
plants is by no means so absolute as to justify the belief of 
their having never been self- fertilised for years. 

Let us now turn to Mr. Darwin's experiments. 

Beseda lutea. The ratio between the heights of the crossed 
plants and those of the self -fertilised were as 100 : 85, the 
weights as 100 : 21, when the plants were grown in pots. 
When grown in open gi'ound they were nearer equality, viz., 
in height, as 100 : 82, and in weight (a better test than 
height), as 100 : 40. Differences in fertility are not given, 
and, therefore, presumably not striking. 

* Fertilisation, etc., p. 116. 

t Mr. Darwin says Papaver vagum, included in the list of plants 
sterile without insect aid, produced a few capsules in the early part 
of the summer; see above, Chap. XXV., on Sexuality and Environment, 


Eeseda odorata. The results of plants grown in pots 
were as follows, the proportions being taken as before. The 
heights were as 100 : 82; weights as 100 : 67; while their 
heights when the plants were grown in the open were as 

He next raised seed bj crossing some flowers and self- 
fertilising others on the same plant of a particular semi- 
self-sterile individual. From these the seedlings gave the 
following results : heights as 100 : 92 ; weights as 100 : 99 ; 
fertility as 100 : 100. 

These results show that the differences have practically- 
vanished ; the weight being a much better test than height, 
as it points to greater assimilative powers, and leaves nothing 
to be desired. 

It is difficult, then, to see how Eeseda furnishes data for 
any argument raised to prove the existence of injuriousness 
in the self-fertilisation of plants. Indeed, Mr. Darwin him- 
self observes : " I expected that the seedlings from this 
semi-self-sterile plant would have profited in a higher degree 

* Mr. Darwin remarks upon this result as follows : " We have here 
the anomalous result of the self-fertilised plants being a little taller 
than the crossed, of which fact I can offer no explanation. It is, of 
course, possible, but not probable, that the labels may have been inter- 
changed by accident " (Cross, etc., p. 121). In my paper (p. 383) referred 
to I have shown that it was most generally the case that while a close 
competition in the same pot proved disadvantageous to the self-fertilised 
seedlings, yet, when they had no competition, the differences were not 
nearly so marked. There are apparently but two alternatives to 
appeal to in order to account for the fact that intercrossed plants are 
not so greatly superior to the self-fertilised when planted in open 
ground, as when in competition in pots ; viz., either the intercrossed 
plants become deteriorated on being planted in open ground, which is 
absurd, or else the self-fertilised must regain or acquire vigour in a 
relatively go-eater degree than do the intercrossed, and thus would seem 
to evince what might be called a greater " elasticity " of growth than 
their intercrossed competitors. 


from a cross than did the seedlings from the fully self-fertile 
plants. But my anticipation was quite wrong, for they 
profited in a less degree : " * — really not at all, for the self- 
fertilised were superior. " An analogous result followed in 
the case of Eschsclioltzia^ in which the offspring of the plants 
of Brazilian parentage (which were partially [said to be 
" absolutely " so, on p. Ill] self-sterile) did not profit more 
from a cross, than did the plants of the far more self-fertile 
English stock." * 

Mr. Darwin commenced his experiments by saying, " This 
plant is remarkable from the crossed seedlings not exceeding 
in height or vigour the self-fertilised. On the other hand, 
a cross greatly increases the productiveness of the flowers on 
the parent-plant, and is sometimes necessary in order that 
they should produce any seed. Moreover, plants thus de- 
rived are themselves much more fertile than those raised 
from self-fertilised flowers ; so that the whole advantage of 
a cross is confined to the reproductive system." t 

Twelve flowers crossed produced eleven good capsules, 
containing 17*4 grains of seeds; eighteen self -fertilised 
flowers produced twelve good capsules, containing 13"61 
grains: therefore the ratio of fertility was as 100:71. In 
the first season the heights were as 100 : 86. Being cut 
down, the next season, they were reversed, " as the self- 
fertilised plants in three out of four pots were now taller 
than and flowered before the crossed plants." 

*' In the second generation,, eleven pairs were raised and 
grown in competition in the usual manner. The two lots 
were nearly equal during their whole growth, or as 100:101. 
There was no great difference in the number of flowers and 
capsules produced by the two lots, when both were left freely 
exposed to the visits of insects." 

* Cross and Self Fertilisation, p. 121. f L.c, p. 109. 



This concludes his experiments with English plants -, and 
though crossing did little or no good, and the first average 
of heights, viz. 100 : 82, he thinks were accidental, the 
converse proposition, that self-fertilisation was injurious, is 
in no way proved. It would be just as logical to say that, 
since the self-fertilised plants grew more vigorously after 
both were cut down, that crossing must have weakened the 
constitution of the crossed seedlings. Or, again, from the 
second year's results, we might justly conclude that the two 
effects were quite identical. 

He next experimented with seed the parents of which 
had been cultivated in Brazil, in which country Fritz 
Miiller had found them to be "absolutely self-sterile with 
pollen from the same plant, but perfectly fertile when ferti- 
lised with pollen from any other plant." Seeds raised from 
these in England " were found not to be so completely self- 
sterile as in Brazil." The average number of seeds produced 
in the capsules borne on the intercrossed and self-fertilised 
plants of Brazilian origin were 80 and 12 respectively in the 
first year; that is in the ratio of 100 : 15. 

With regard to the second generation, or grandchildren, 
next raised, Mr. Darwin observes : *' As the grandparents ia 
Brazil absolutely required cross-fertilisation in order to 
yield any seeds, I expected that self- fertilisation would have 
proved very injurious to these seedlings, and that the crossed 
ones would have been greatly superior in height and vigour 
to those raised from the self-fertilised flowers. But the 
result showed that my anticipation was erroneous ; for as in 
the last experiment with plants of the English stock, so in 
the present one, the self-fertilised plants exceeded the crossed 
by a little in height, viz., as 100 : 101." 

In the next year the average number of seeds per capsule 
of the crossed and self -fertilised was as 100 : 86*6 ; so that the 


relative fertility of the self-fertilised had risen from zero in 
Brazil to 15, and then to 866 pc, in comparison with the 
crossed regarded as 100. 

He now made crossings between the offspring of the 
Brazilian plants and the English-grown plants, with the 
following results : — 

First, as to heights, — 

The English-crossed to the self -fertilised plants 100:109 

The English-crossed to the intercrossed * plants ... ... 100: 94 

The intercrossed to the self -fertilised plants ... ... ... 300:116 

Secondly, as to weights, — 

The English-crossed to the self -fertilised plants 100 : 118 

The English-crossed to the intercrossed plants ... ... 100:100 

The intercrossed to the self -fertilised plants 100:118 

Three rows of plants of each kind grew in the open ; and 
here also the self-fertilised grew taller than the others. 
Moreover, all except three of the self -fertilised were killed 
by the winter. 

" We thus see that the self -fertilised plants which were 
grown in the nine pots were superior in height (as 116 : 100) 
and in weight (as 118 : 100), and apparently in hardiness, to 
the intercrossed plants derived from a cross between the 
grandchildren of the Brazilian stock. The superiority is 
here much more strongly marked than in the second trial 
with the plants of the English stock, in which the self- 
fertilised were to the crossed in height as 101 : 100. It is 
a far more remarkable fact . . . that the self- fertilised plants 
exceeded in height (as 109:100), and in weight (as 118 : 100), 
the offspring of the Brazilian stock crossed by the English 

* "Intercrossed" signifies the ofifspring of the Brazilian plants 
crossed with one another. 


When we look back and remember that the plant was 
" absolutely self-sterile " in Brazil, and compare that fact 
Avitli these final results, it is difficult to see how self-fertilisa- 
tion can be charged in any way with injurionsness. Though 
the results may have shown little or no advantage from 
crossing, it does not follow " that the differences," namely 
greater height, weight, or fertility of the self-fertilised, were 
attributable " to the inferiority of the self-fertilised seedlings, 
due to the injurious eifects of self -fertilisation." 

On the other hand, the facts appear to warrant the 
conclusion that this north-temperate plant became barren 
in Brazil in consequence of the hot climate ; that the 
recovery of its self -fertilising powers was due to the English 
climate better suiting it; that it at once responded to the 
effort, so that its self- fertility rose in two generations from 

to 86-6 p.c. The plants, too, thus raised showed nothing to 
indicate any constitutional derangement that might, with 
any show of reason, be attributable to self-fertilisation. 

From the preceding observations upon Mr. Darwin's 
reasoning, I think the reader will now see that it is not so 
conclusive in proving the existence of any injurionsness in 
self-fertilisation as he appeared to think. 

This chapter was already in type when I met with the 
following passage in " The Life and Letters of C. Darwin," 
written in May, 1881 : " I now believe . . . that I ought to 
have insisted more strongly than I did on the many adap- 
tations for self -fertilisation, though I was well aware of 
many such adaptations." 

With reo-ard to the values of other kinds of fertilisation, 

1 must refer to Mr. Darwin's works; for it is beyond my 
purpose to discuss them, as they have no special bearing 
upon the origin of floral structures. 



The Origin of Species by Insect Agency. — The attractive 
features of flowers being now well recognized as correlated 
with insect agency in fertilisation, the question arises, How 
have they come into existence ? We may suppose that a plant 
bore seedlings, some of which had, we will say, the corolla 
accidentally (that means from some unknown cause arising 
from ivifJiiu) larger on one side than another ; and then such 
a flower, being selected by insects, left offspring which, by 
gradual improvement through repeated selection, ultimately 
reached the form it now possesses. 

As an alternative, we may suppose that the first impulse 
came from without, and induced by the insect itself; so that 
the variation once set up in a definite direction, went on 
improving under the constantly repeated stimulus of insect 
visitors until the form of the flower was actually con- 
formable to the insect itself. 

The process of evolutionary development might perhaps 
be much the same under either supposition, but the latter 
hypothesis has more than one advantage. First, in the 
assignment of a direct physical cause for the incipient change, 
instead of some incidental and unaccountable variation, 
which must be assumed by the former. Secondly, the theory 
does not require the plant to make an indefinite number of 


less useful clianges or variations, only to be discarded at 
eacli generation for the one form that was wanted. Thirdly, 
as a great number of flo^yers would be visited, both on one 
plant and on many surrounding individuals in the neigh- 
bourhood, great numbers might bear offspring advancing 
more or less in the same direction ; and there would be no 
fear of extermination, even if some happened tb be crossed 
by the parent form. Indeed, the varying offspring would 
largely supersede the parent form in number altogether, if 
they sprang up at one place without emigration. If we 
supply the additional aid of isolation, many otter influences 
would be brought to bear upon them, and they would be free 
to vary without any interference from the parent stock. 

Mr. Darwin has abundantly shown that when a plant is 
crossed, and its seedlings strugi^le in a confined place with 
those derived from flowers which, have not been crossed but 
artificially self-fertilised, they generally succeed in mastering 
the latter ; so that if there be any struggle with the seedlings 
of a self-fertilised parent, such a struggle for life is mainly 
during the early period of growth, before any varietal or 
specific characters of the flowers have put in an appearance 
at all. For it is only in the youthful stages that the greatest 
contest is maintained ; and the result depends largely upon 
constitutional, and not at all upon specific, that is morpho- 
logical characters, mostly taken from the flowers. Now, 
Mr. Darwin has shown that such constitutional vigour does 
very generally accompany at least the first few years of 
crossing. So that we have a vera causa of the success 
of such newly crossed plants in the preliminary struggle 
for life. It need hardly be remarked that if insects thus 
start a new variety, they are crossing the flowers at the 
same time. 

It is true that the stimulus of crossing does not last for 


many years ; but it is probably all tliat is wanted to give 
the crossed plant the ascendancy wlien starting on an evolu- 
tionary career. 

As an illustrative case of a struggle between two varieties, 
I took the same quantity of English-grown " Revett's " 
wheat and Russian " Kubanka," the former having a pre- 
ponderance of starch and the latter of gluten, being a smaller 
and harder grain. I sowed them as thickly as possible on a 
square yard, the two kinds having been previously well 
mixed together. They all germinated, and the struggle of 
course became intense. About twenty ears only were pro- 
duced, which were all Kubanka. The experiment was 
repeated a second year, with the same result. This is what 
I would consider as, therefore, due to " constitutional 

Survivors, however, are by no means entirely dependent 
upon constitutiou, much less on specific differences ; for seeds 
which fall on the circumference of the crowd, or on a better 
soil than that upon which others may happen to lie, as on 
stony ground, are thereby " selected," but it is through no 
merit of their own, as in any way being the fittest, for they 
survive only because they are the " luckiest ; " just as out of 
the thousands of eggs of a salmon a few only escape the 
jaws of their enemies : so that simply " good luck " plays an 
important part in determining which shall survive and come 
to maturity in both kingdoms alike. 

Hence, during the period of life when the struggle for 
existence is most intense, there are various circumstances 
which determine what plants shall survive ; and in probably 
few cases, generally no case, have the morphological variations 
or specific characters any voice in the matter of selection 
whatever, excepting indirectly, as stated above, whenever 
constitutional vigour is correlated with first crossings. 


The difficulty which Mr. Romanes has felt in the struggle 
for life through the swamping effect of a varying offspring 
being crossed with the parent form, seems to me to be illusory 
as far as most flowering plants are concerned.* For not only 
do the majority of new forms arise through transport of 
seeds to a new and distant locality, but even at home, if the 
plant be at all responsive, so many seedlings, perhaps all, 
will tend to be difiFerentiated at the same time and in the 
same way, that the parent form will soon be in a minority, 
and if now neglected by insects may die out through " insect- 
selection " of the new form. 

According to the old view, that plants are varying spon- 
taneously in all directions, and that only a few are selected 
by insects, the difficulty has long been felt that dangers of 
all sorts must surround the offspring of those few. Let us 
reverse the process, however, and let the insects themselves 
be the cause of changes set up in the flowers in the adaptive 
directions, and the responsive power of the flower itself will 
soon develop the best forms. These run no risk of being lost, 
through the multitude of offspring. Hence, if my theory be 
true, physiological selection, which I cannot find horticul- 
turists are inclined to accept, is not needed at all. 

Suppose some prevailing insect to have begun to set up 
incipient changes for a new variety, which then becomes dis- 
persed ; since many of the offspring will possess the new- 
adaptation, and several other kinds of insects will visit the 
flower in different places, as the seeds happen to get trans- 
ported, the result will be, that while the original species 
of insect induces the descendants of the plant at home to 
vary in adaptation to itself, others are at work elsewhere, 

* Fritz Miiller found the genus Abutilon and a species of Bignonia to 
be more or less sterile with parental pollen. See Miiller's Fertilisation, 
etc., pp. 145, 466. 


modifying the same incipient alterations to suit themselves. 
Hence, as soon as isolation by migration has taken place, it 
is the presence of other insects which determines the develop- 
ment of other varieties. All, however, are based on the 
same plan of departure. 

In this way many varietal and subsequently specific forms 
of the same genus will arise; and the further they travel 
from the parental home the greater, perhaps, Avill be the 
specific differences; and thus can representative species be 
accounted for, especially among conspicuously flowering 

On the other hand, the perpetually self-fertilising species 
which alone, as a rule, are cosmopolitan, are almost identical 
in form, or at least have a minimum of differences between 
them, and such as may possibly be accounted for by climatal 
causes alone. 

Difficulties of Natural Selection. — The greatest diffi- 
culty I have always felt in the idea that a plant was selected 
because it had some floral structures more appropriate than 
others, lay first in the fact that the principal period of the 
struggle for life takes place in the seedling stage, before any 
varietal and specific characters have appeared ; and, unless 
there w^ere a large number of the seedlings which would 
ultimately bear the improved flower, or else a superior con- 
stitutional vigour be guaranteed to be correlated with the 
particular varietal characters to be preserved, these alone 
could have nothing to do with the survival of the fittest. 

Secondly, granting that the plant has succeeded in sur- 
viving till the flowering period, then why should so many 
minute details of floral structure be nexessarily correlated ? 
If the loss of three out of five carpels in the Lahiatce w^ere 
due to natural selection, why should this go hand-in-hand 
with a multiplication of the ribs of the calyx, and the 


peculiar lipped and hooded corolla with the lateral position 
of the flower, etc. ? We find in selecting peas and beans 
great varieties among them, but next to none in the calyx 
and corolla, to which the horticulturist pays no attention. 

In nature, however, we often find in flowers regularly 
visited by insects innumerable and minutely correlated adap- 
tations in all the whorls, which must have all varied together 
to form such existing flowers. Now, the difficulty of their 
doing so without some common cause, which affects them 
all simultaneously, seems to me insuperable. 

If my theory, however, be accepted, it solves the whole 
mystery at once, as all the changes are set up by one prime 
cause, namely, the irritations of the insect in the case of all 
flowers adapted to insect-fertilisation ; while the absence of 
insects in regularly self-fertilised flowers, as well as anemo- 
philous ones, is sufficient to account for the atrophy which 
has affected them, the present condition of such flowers 
having been the inevitable result. 

Hence, instead of speaking of the Origin of Species of 
Plants by Natural Selection, I would regard the survival of 
the fittest as first issuing from " Constitutional Selection ; " * 
while the origin of the floral specific characters is the result 
of the responsive power of protoplasm to external stimuli. 
These latter are infinitely various in kind and degree, as has 
been shown in the early part of this book. The result is, that 
while high differentiations occur in some directions, degrada- 
tions are met with in others, sometimes seen in different parts 
of the entire plant ; but not at all infrequently are both 
features observable simultaneously in one and the same floral 

* Of course the chances of less competition by growing on the circum- 
ference of the batch of seedlings, by receiving a little more light, etc., 
aid in selecting, and sometimes may determine, as stated above, those 
which shall survive. 


"wtorl. The phrase " natural selection " will therefore have 
been noticed as conspicuous by its absence throughout this 
book. This is not because I would in the least deny the 
fact that vast numbers of seedlings perish while others sur- 
vive through that form which I have called " constitutional 
selection," which are thus " selected," and arrive at the 
flowering and fruiting stages; and, again, that of these latter 
many may set no seed through the neglect of insects, etc., and 
so perish entirely and leave no offspring, while others again 
survive and are selected. Why, however, I do not refer any 
particular structure to the action of natural selection is 
because I have always felt or perceived a danger in doing so. 
Natural selection is, as thus styled, an abstraction ; and as 
long as we hide our ignorance of its concrete representatives, 
that is to say, the real causes at work to induce a change, we 
may fancy we understand all about it, while we may be in 
reality in profound ignorance. 

Professor Huxley remarked, in his lectures on the Origin 
of Species, that what we want is " a good theory of varia- 
tion." It is in the attempt to fill this hiatus that I have, 
step by step throughout this book, preferred to give what 
seemed to me a direct cause, mechanical, physiological, 
climatal, etc., for every structure ; which may bring us 
nearer to a comprehension of the direct interaction of cause 
and effect than the vague term " natural selection " seems 
capable of doing Thus, to take an example, Miiller refers 
the loss of the fifth stamen in Labiates to natural selection, 
but makes no statement how he supposes selection to have 
done it. On the other hand, I would prefer to attribute its 
absence to atrophy, in compensation with the hypertrophy 
of the corolla on the posterior side. I may be wrong, of 
course, but at all events I give a reasonable cause, which is 
a fertile one in bringing about alterations in the structure 


of flowers ; whereas " natural selection " leaves us exactly 
where we were before. Moreover, natural selection is made 
to cover exactly opposite processes ; for the formation of the 
enlarged lip, on the one hand, would be attributed to it, just 
as much as the elimination of a stamen altogether, on the 
other. Instead, therefore, of using this term as the cause of 
anything and everything, I prefer to attribute effects to 
hypertrophy, atrophy, resistance to strains, responsive action 
to irritations, and so on. If it be thought that natural selec- 
tion somehow underlies all this, the reader is at liberty 
to substitute the phrase; but, I must confess, it conveys 
nothing definite to my mind, while the others undoubtedly do. 
I do not wish the reader to suppose that my theory is 
altogether in opposition to Mr. Darwin's ; for it must not be 
forgotten that he himself laid great stress on the environ- 
ment as a cause of variability upon which, when once brought 
about, natural selection could then act. Thus he remarks : 
" To sum up on the origin of our domestic races of animals 
and plants. Changed conditions of life are of the highest 
importance in causing variability, both by acting directly on 
the organisation, and indirectly by affecting the reproductive 
system. It is not probable that variability is an inherent 
and necessary contingent, under all circumstances. . . . Vari- 
ability is governed by many unknown laws, of which corre- 
lated growth is probably the most important. Something, 
but how much we do not know, may be attributed to the 
definite action of the conditions of life. [Under this I would 
include the definite action of insects exerted mechanically 
npon the organs of flowers.] Some, perhaps a great, effect 
may be attributed to the increased use or disuse of parts. 
[Compensation plays undoubtedly a very important part]. . . 
Over all these causes of Change, the accumulative action 
of Selection, whether applied methodically and quickly, or 


unconsciously and slowly, but more efficiently, seems to have 
been the predominant Power." * 

If thus the variations of floral structures can be reasonably 
referred directly to external agencies, and we may speak of each 
as a cause instead of using the abstract expresssion " natural 
selection," there still remains the question. What has brought 
into existence the primary flowers themselves, which insects 
have subsequently modified into their present conditions ? 

The Origin of Flowers. — There are good reasons for 
regarding Gymnosperms — both from their extreme antiquity, 
as well as from points of structure showing affinity with 
the higher Cryptogams ; such, for example, as the Lycojpodi- 
aceoe — as standing in some sort of way intermediate between 
the latter and Dicotyledons. Yet the connecting links are 
much wanted on both sides of them. As far as Coniferce 
and Cycadem can help us, we are strongly led to believe that 
they were primitively, just as they are now, anemophilous 
and diclinous ; though the subdioecious (?) WehvitscMa has 
points of structure which seem to indicate its being a 
degraded state of an hermaphrodite plant. This remarkable 
monotypic genus is, however, too isolated and unique to afford 
any safe point of departure on the road to Dicotyledons, so 
that with regard to the latter we are still driven to specu- 
lation alone. 

If, then, we are right in assuming Gymnosperms to 
have been always diclinous, and Dicotyledons to have arisen 
from some member of that group, then it is presumable that 
the first were diclinous, perhaps dioecious, and anemophilous 
as well. The general opinion seems to be that they were 
dioecious; and Mr. Darwin thought that moncecism was the 
next step, and thence hermaphroditism was ultimately 

* Origin of Species, 6th ed., p. 31. 


Now, Tve must not forget that when a female flower is 
pollinated the effect of the impregnation by the pollen-tube 
is not only to create an embryo in the ovule, but to endow it 
potentially with its own sexuality ; so that the sexless embryo 
becomes potentially both male and female ; in as much as it 
may subsequently grow up to be solely a male or solely a 
female plant ; or else it may combine the sexes, either as a 
monoecious or hermaphrodite plant. 

Moreover, we now know that the resulting sex which 
appears in dioecious plants on maturity is largely, if not 
entirely, dependent upon conditions of nutrition, possibly 
aided by other and unknown influences. 

Consequently, we cannot say for certain whether the 
first Dicotyledons were nob at least monoecious, if not 
hermaphrodite, since the former of these states prevails 
already in Gymnosperms, as in Finns; while the latter is 
hinted at in not infrequent monstrous conditions when the 
lowermost scales of the spiral series in cones of Abies excelsa, 
etc., are antheriferous, instead of being ovuliferous. * Such 
cases show that one (the male) sex can suddenly appear in 
the same spiral series as the other. And this is all that is 
wanted to form an hermaphrodite flower; for continuously 
spirally- arranged sexual organs are characteristic of many 
plants, such as of the Ranunculacece ; and such a monstrous 
condition may simply be a reversion to a primitive her- 
maphrodite state. Hence appears the inherent possibility 
of the production of hermaphroditism without any slow 
evolutionary process at all ; but simply as a result of the 
conveyance of the male energy to the female plant, by the 
very act of pollination itself. 

Mr. Darwin, when speculating on the origin of herma- 
phroditism, wrote as follows : " By what graduated steps 
* Teratology, p. 192. 


an hermaplirodite condition was acquired we do not "know. 
But we can see that if a lowly organised form, in whicli the 
two sexes were represented by somewhat different individuals, 
were to increase by budding either before or after conjugation, 
the two incipient sexes would be capable of appearing by 
buds on the same stock, as occasionally occurs with various 
characters at the present day. The organism would then be 
in a monoecious condition, and this is probably the first 
step towards hermaphroditism ; for if very simple male and 
female flowers on the same stock, each consisting of a single 
stamen or pistil, were brought close together and surrounded 
by a common envelope, in nearly the same manner as with 
the florets of the Comjoositce, we should have a hermaphrodite 
flower." * 

It is a singular fact that !Mr. Darwin never seems to have 
thought of Euphorhia, which tallies exactly with his hypo- 
thetical origin of a hermaphrodite flower ; but, unfortunately, 
a " blossom " of an Eujplwrhia is not regarded by botanists 
as a flower, but an inflorescence. It consists of a " single 
pistil," on its own pedicel, surrounded by many " single 
stamens," each on their own pedicels ; and are " brought 
close together and surrounded by a common envelope." 

Mr. Darwin's mistake resides in his supposition that 
hermaphroditism must have arisen from dioecism, by passing 
through monoecism ; so that he is obliged by this order of 
progress to consider a flower with stamens and a pistil to be 
made of separate flower-buds, i.e. to be axial structures with 
their appendages reduced to at least one of each kind. But 
from phyllotactical reasons, it is clear that the origin and 
arrangements of the floral members are entirely foliar. 

All that seems necessary for us to assume as the origin of 
a flower with a conspicuous corolla or perianth, is a leaf -bud 
* Cross and Self Fertilisation of Plants, p. 410. 


of which some of the members have already differentiated 
into carpellary, others into staminal organs, the enter 
appendages being simply bracts, like, we will say, those 
surrounding the stamens or ovule of the Yew. 

As insects often come for pollen alone — as in honeyless 
flowers of Laburnum, Poppies, St. John's Wort, and Roses, — 
and then pierce the juicy tissues for moistening the honey, as 
they have been seen to do in Anemone, Laburnum, Hyacinths, 
Orchis^ etc., we may, I think, infer with some probability 
that they did the same with the primitive flowers. 

Having once attracted insects to come regularly, then a 
multitudinous series of differentiations would follow. The 
corolla would in all probability be the first to issue out of the 
bracts, as being the next whorl to the stamens and as a 
result of stimulus ; other changes, already described under 
the Principles of Variation, would follow by degrees and in 
different combinations ; but in every case they would be due 
to the responsive action of the protoplasm in consequence of 
the irritations set up by the weights, pressures, thrusts, 
tensions, etc., of the insect visitors. 

Thus, then, do I believe that the whole Floral World has 


Adelphous filaments, 57; imitated, 
59 ; and nectaries, 58 

Adhesion, analogies in animal king- 
dom of, 48, 88 ; principle of, 5, 78, 
seqq, ; rationale of, 80 ; of stamen to 
perianth, and origin of, 81, and to 
style (?), Aristolochia, (fig. 21) 83 

^stiv^ations and phyllotaxis, (fig. 3) 

Alpine, flowers, colours of, 176 ; 
strawberry, phyllody of, 301 

Amaryllis, appendage to perianth, 
(fig. 41) 134 

Andiodicecism, examples, explanation 
and origin of, 227 

Andrcecium, explained, 4 ; irregu- 
larity in, origin of, 109 

Anemophilous flowers, 265, seqq. ; 
characters of, 268 ; cosmopolitan, 
283 ; " long-lived " stigmas of, 
269 ; pollen of, 266 

Anemophily, and Greenland flora, 
270 ; and cleistogamy, 264 ; and 
degeneracy, 266, 272 ; and hete- 
rogamy, 269 ; origin of, 266, 270, 
272; and protogyny, 200, 269, 

Anisomerous whorls, explained, 5 ; 
causes of disarrangement of, 45 

Anthers, on bracts, (fig. 64) 288; 
connivent, of Violet, 60 ; conta- 
bescent, 275 ; on glumes, (fig. 65) 
288 ; metamorphosed, 293, (fig. 81) 

298, (figs. 83, 84) 302; stigma- 
tiferous, (fig. 76) 294 ; syngene- 
sious, and interpretation of, (fig. 
11) 60; versatile, 266, 268 

Ant-plants, hereditary effects of irri- 
tation in, 115, 142,' 157 

Appendages, in Amaryllis, (fig. 41) 
134 ; and axis, homology between, 
309 ; origin of floral, 133 

Aquilegia vulgaris, arrangement of 
floral whorls of, 22 ; number of 
parts in, 22 

Arahis alhida, leaf-traces of, (fig. 7) 

Arctic flora, and anemophily, 270 ; 
and self-fertilisation, 259 

Aristolochia, structure of flower, (fig. 

Arrangement, causes of, 47 ; displace- 
ment of, by anisomery, and substi- 
tution, 45 ; illustrations of, in 
BanunculacecB, 21, seqq.; principle 
of, 5, 139 

Arrest, of carpels, 4, 8, 278; of 
carpels in Campanitlacece, 44; of 
floral axis, 6 ; in free-central pla- 
centas, 72, seqq. ; of growth of 
ovary and seeds in Orchids, 169, 
281, and in Willows, 170 

Atragene, staminal nectaries of, (fig. 
44) 141 

Atrophy and hypertrophy in animal 
kingdom, 88 ; as causes of irregu- 
larities, 108 ; in compensation, 105 ; 
in zygomorphism, 116, seqq. 



Autogamy, explained, 198, 311. See 

Axis, and appendage, homology be- 
tween, 309 ; floral, cause of arrest 
of, 6 


Beta^ formation of ovule of, (fig. 16) 

Boughs, curvature of, due to strain, 

(fig. 39) 125 
Bracts, petaloid, 286, (figs. 62, 63) 

2h7 ; pistiloid (glumes), (fig. 65) 

288 ; progressive changes in, 286 ; 

transitional forms of, in Hellebore, 

(fig. 61) 286 
Bulbs, origin of, from funicle, 310 ; 

from leaf-sheath, 310 

Cabbages, excrescences on, homologous 
with ovules, 307 

Calyx, arrest of, 8, 184, 194; pro- 
gressive metamorphosis of, 288 ^ 
-tube, 89, seqq. See Sepals. 

Campanula medium, anatomy of flower 
of, (fig. 8)43, (fig. 15)71 

CampanulacecE, arrangement of carpels 
in genera of, 44 

Capparidece, andrcecium of, and sym- 
metry in flower of, 33 

Carpels, arrest of, 4, 8, 278 ; in Cam- 
panulacece, 44; cohesion of, 62; 
decrease by compensation, 21, 278 , 
phyllody of, 302 ; superposition of, 
44, seqq. ; typical num.ber of whorls 
of, 4. See Pistil. 

Carpophore, placental origin of, 72 

Cell-division and light, 154 

Cell-wall, thickening of, to resist 
pressure, 127 

Centaurea, adaptations for fertilisa- 
tion, (fig. 11) 60 ; and sexuality, 

Change of symmetry, 18, 186 

Chorisis, and arrangement, 24, 39, 44, 

46 ; multiplication of stamens by, 
44, and of carpels by, 44, 308, and 
of ovules by (in Orchids), 309 

Cleistogamy, and anemophily, 264 ; 
and degeneracy, 251, seqq. ; and en- 
vironment, 263 ; explained, 198; in 
flowers, 251 ; illustrations of 257- 
262; in Impatiens, (fig. 58) 261; 
in Lamium, (fig. 59) 261 ; origin 
of, 262-264; in Oxalis, (fig. 57) 
260; in Salvia, (fig. 60) 262; in 
Violets, (figs. 55, 56) 257, 258 

Cohesion, of carpels, 62 ; illustrations 
of, 49, 50; origin of, 50 ; of petals, 
56 , in Fhytewm, (fig. 9) 50 ; 
principle of, 5, 48 , of sepals, 54 , 
of stamens, 57 , to resist strains, 
51, 53 ; varieties of, congenital and 
by contact, 48 

Colours, of Alpine flowers, 176 ; 
changes in, 176 ; and darkness, 177 ; 
effect of crossing on, 178 ; eflect of 
salts on, 175 ; of flowers, 174 , and 
insects, 182; laws of, 174; nutri- 
tion and, 178 ; origin of, 178 ; as 
pathfinders, 178, and arrest of, 
253 , white and pale tints, and 
self-fertilisation, 253; whole, and 
self-fertilisation, 183 

Compensation, in adaptations of 
flowers, 105, 117 ; atrophy and 
hypertrophy in, 105 ; increase of 
seeds and decrease of carpels by, 
21, 278 ; in irregular flowers, 103, 
seqq. ; in rudimentary organs, 284 

Conducting tissue, of Orchids, 165; 
origin of, by irritation of pollen- 
tube, 165, seqq. ; structure of, (fig. 
50) 164 

Coniferoe, foliage of, adnate and free, 
84; origin of flowers and the, 337 

Connivent anthers, of Violet, 60 

Contabescence of anthers, 275 

Cords, fibro-vascular, alteration in 
orientation of, 64, 65 ; as floral 
I units, 300, 308, 309 ; in flower of 
Campanula, (fig. 8) 43, (fig. 15) 
71 ; increase in number of, 55-57 ; 
orientation of phloem and tracheae 



in, 63 ; in receptacular tubes, (fig. 
U) 68, (fig. 28) 95, (fig. 30) 97 ; 
sepaline, of Salvia, 55 ; as origiu 
of the staminal and carpellavy, in 
MalvacecB, 43, 44 

Corollas, appendages to, origin of, 133, 
seqq. ; form of, 101, seqq. ; meta- 
morphoses of, 292, 301 ; movements 
in, of Genista, (fig. 47) 160 ; of 
Lopezia, (fig. 48) 161 ; origin of, 
irregular, 103, seqq.; petals of, 
displacement of, by insects, (figs. 
33-35) 110, 111; polliniferous, 
292, 293 ; progressive metamor- 
phoses of, 292 ; reduction of size 
or,9, 254, in Geranium, 252; regular 
and irregular, 101, seqq. ; sensi- 
tiveness in, Ypoma:a,lQ\; stameni- 
ferous, (figs. 72, 73) 292, 293; 
strains, effect of, on the formation 
of, 101, seqq., 126 ; structure of 
bilateral, 116, seqq. ; virescence of, 
(figs. 83, 84) 301, seqq. See Petals. 

Correlation of growth, 112, 113, 
117 ; irregularities by, 108 

Cross-fertilisation, advantages of, in 
evolution of species, 330, and in 
horticulture, 311 ; colour, effects 
on, 178 ; disadvantages of, 314 ; 
rationale of, 312 ; stimulus pro- 
duced by, 312 ; views of Mr. Darwin 
on, 315 

Cruciferoe, anatomy of floral recep- 
tacle, (fig. 6) 32 ; symmetry of, 


Darkness and colours, 177 

Declinate stamens, in Dictamnus, (fig. 
33) 110; distribution of forces in, 
of Echium, (fig. 20) 82 ; of Epilo- 
bium, (fig. 34) 111; origin of, due 
to weight of insects, 110, 111 

Degeneracy and degradation, of an- 
droecium, 273 ; and androdicecism, 
227 ; and anemophily, 266 ; of 
flowers, 251, seqq.; in inconspicuous 
flowers, cause of, 251 ; in Orchids. 

172, 281, 319 ; origin of, 282 ; and 

self- fertilisation, 252, seqq. 
Development, of floral whorls, 191, and 

continuous during flowering, 122 ; 

order of, of parts of flowers, relative 

only, 195; rates of, in pistil, 192, 

Dialysis, explained, 5, 50 ; in Mimu- 

lus, (rig. 10) 51 
Diclinism, and heterostylism, 228 ; 

partial, 220 ; in primitive flowers, 

Dimorphism, and fertilisation in Viola 

tricolor, 255 ; and heterostylism, 

203 ; in stamens, (fig. 37) 121 
Dicecism, and heterostylism as cause 

of, 218; of primitive flowers, 337 
Domatia, hereditary formation of, 

115, 142, 157 
Doubling, causes of, 298 
Drosera, metamorphoses of tentacles 

of, into ovules, 307 
Duvernoia, zygomorphism of, origin 

of, (fig. 31) 107 


Electricity, effects on protoplasm, (fig. 

45) 152, on nucleus, 154 
Emergence, alteration in order of, in 

regular and in irregular flowers, 

187 ; and development of ovules, 

195, and interpretation of, 196; of 

floral whorls, 184; order of, 184 
Energy, reproductive and vegetative, 

231, seqq. 
Environment, action of, Mr. Darwin's 

views on, 336; influence of, 158; 

origin of species through, 329, 

seqq. See Preface. 
Epidermis, origin of root hairs on, 

(fig. 42), 137 
Eranthis, arrangement and number of 

parts in flower of, 22 
Exclusion, of insects from flowers, 

102, 133, seqq. 
Excrescences, on corolla, (fig. 87) 

306 ; on cabbage-leaves, as homo- 

logues of ovules, 307 



Fasciation, 51, 85 ; of petioles of pear, 
(fig. 26) 94 

Fertilisation, cross- (see s.v.) ; and 
origin of species, 329 ; by pollen- 
tube (see s.v.) ; varieties of, 311; 
self- (see s.v.) 

Fibro-vascular cord, as a fundamental 
unit, 300, 308, 309. See Cord. 

Flora, of Dovrefjeld, and self-fertilisa- 
tion, 259 ; of Galapagos Islands, 
270 ; of Greenland, 270 

Floral symmetry, correlation with 
phyllotaxis, 14 ; explained, 4, 5 ; 
variations in, 12 

Florai whorls, development of, order 
of, 191; emergence of, 184; sym- 
metrical decrease and increase in, 
18 ; unsymmetrical, 20. See Whorls. 

Flowers, conspicuous, development of 
parts of, 191; degeneracy in, 251 ; 
inconspicuous, origin of, 251 ; origin 
of, 337 ; typical, structure of, (fig. 

Forces, effects of mechanical, etc. 
See Mechanical forces. 

Forms, of floral organs, 101, seqq. ; 
dimorphic, of stamens, (fig. 37) 
121; principle of, 5; transitional, 
118, seqq. 

Funicle, bulb arising from, 310 ; as 
origin of ovule, 303" 

Galls, analogous to tumours, 144 ; 

due to irritation, 144 ; hairs of, 

Garidella, arrangement and number 

of parts in, (Hg. 4) 21 
Glands and rudimentary organs, 283. 

See Nectaries. 
Growth of organs, continuous during 

flowering period, 122 ; correlation 

of, 112, 333 
Guides, degeneracy of, in self-fertil- 
ised flowers, 253 ; origin of, 178 

Gymnosperms, and the origin of 
flowers, 337 

Gynandrous, 82 ; Aristolochia, (fig. 

Gynodicecism, causes of, 221, seqq. ; 
and climate, 221 ; explained, 220 ; 
origin of, 222, seqq. ; and soil, 221 

Gyncecium, degeneracy of, 278 ; ex- 
plained, 4 ; unsymmetrical decrease 
in, 20. See Carpels and Pistil. 

Gynomoncecism, examples of, 226 ; 
explained, 220 

flairs, on filaments, origin of, 136 
(see fig. 11, 60); in galls, 138; oa 
roots, origin of, 137 ; on seeds, 
170 ; within styles, origin of, 139 ; 
tangles and wheels, origin of, 133, 

Heliotrope, stigma of, cause of ano- 
malous, 135 

Hellebore, alteration in orientation of 
cords, (fig, 12) 64; arrangement 
and number of parts of floral 
whorls (fig. 5), 22 

Hercogamy, explained, 317 ; in 
Orchids, 314; relative character 
of, 319 

Hermaphroditism, origin of, Mr. 
Darwin's theory of, and observa- 
tions on, 339 

Heterogamy, explained, 198 ; and 
sexuality, 243 

Heteromorphic flowers explained, 

Heterostylism, explained, 203 ; and 
diclinism, 228 ; and dicecism, 218 
and degrees of fertility, 204, seqq. 
origin of, 213; and sexuality, 244 
structure of stigmas in, 216; un- 
stable, in stamens of Narcissus 
cernuus, (fig. 37) 121 

Homogamy, explained, 198 ; and 
anemophily, 269; fluctuating con- 
ditions about, 201 

Homology, of appendages and axis, 



309 ; explained, 285 ; origin of, 

Honiomorphic conditions, 203 
Homostyled, flowers, explained, 203 ; 

forms of Auricula, 208 ; of Frimula 

Sinensis, 209 
Hooks of Uncaria, (fig. 46) 156 
Hypertrophy, in animal kingdom, 88 ; 

cause of, 51, 88 ; effects of, in 

unions, 86, 87 ; form, a cause of, 

105, seqq.; 116, seqq. ; in Orchids, 

87 ; of placentas, 307 

Illegitimate, or homomorphic unions 

Jmpatiens, secretive stipules of, (fig. 
43) 140 

Impregnation, a form of nutrition, 

Inconspicuous flowers, 251, seqq. ; 
anemophilous, 265 ; due to degene- 
racy, 251, seqq. ; origin of, 282 ; 
self-fertilising, 253, seqq. 

Insects, origin of species by agency 
of, 329 ; relative proportion of, in 
regular and irregular flowers, 102, 
103, 314; visitors to Compositce, 

Irregularity, origin of, 103 

Irritability. See Ant-plants, Appen- 
dages, Form, Protoplasm, Zygo- 

Laws, of alternation, 41 ; of colour, 
174 ; of superposition, 41 

Leaf, cabbage, excrescences on, 307 ; 
of Coniferce, adnate and free, 84, 
85 ; opposite and verticillate, 9 ; 
transition from opposite to verti- 
cillate, (fig. 2) 11, 17, 18. See 

Leaf-traces, of Arabis albida, (fig. 7) 
39 ; compared with floral, 40 

Liber-fibre, origin of, 250 

Light, and cell-division, 154 ; in- 
fluences of, on leaves, 154; on 
roots of Ivy, 155 ; on nucleus, 154 ; 
and sleep of calyx and corolla, 155 

Lysirnachia, anatomy of floral re- 
ceptacle of, (fig. 19) 77 


Mechanical forces, action on boughs, 
125 ; on corolla, 126 ; on pear 
growth, 124; on stamens, 81, 82, 
126 ; tissues, formation of, by, 155, 
seqq. See Irritability. 

Metamorphosis, of bracts, 286 ; of 
calyx, 288 ; of corolla, 292, 302 ; 
of flowers, 285, 295 ; of pistil, 295 ; 
of stamens, 292, 298 ; of tentacles 
o£ Drosera, 307 

Movements, in corolla, 160; of fila- 
ments, 159, 161, 162; of pistil, 
162; of stamens, 162; of staminode, 
161 ; of stigmas and styles, 159, 


Narcissus cernuus, unstable hetero- 
stylism of, 121 

Natural selection, difficulties of, 333 ; 
forms of, 330, seqq. ; insufficient 
as a cause, 335. See Selection. 

Nectaries, 140, seqq. ; and adelphous 
stamens, 58 ; irritation, an origin 
of, 141, 143 ; and pollination, 148 ; 
position of, 140, seqq. ; stamiual in 
Atragene, (fig. 44) 141 ; stipular 
in Impatiens, (fig. 43) 140 

Nepenthes, origin of pitcher of, 146 

Nucleus, effect of electrical irritation 
on, 154, of light on, 154 ; of pollen- 
tube, effect of, 250 

Numbers, illustrations of special, 25- 
38 ; origin of, 9 ; principle of 4, 7 



Obdiplostemony, 188 ; cause of, 190 ; 
origin of, 150 

Opposite and verticillate leaves, 9 ; 
as origin of alternate, 11 

Orchids, adhesive roots of, (fig. 42) 
137 ; conducting tissue of, 165 ; de- 
generacy in, 172, 280, 319; effect 
of irritations on, mechanical, 114, 
of larvae, 171, physiological, of 
pollen-tubes, 165, seqq. ; hyper- 
trophy in, 87 ; monstrous, 87 ; self- 
fertilising, 253, 318 

Order of development of floral whorls, 
relative only, 195 

Organs, floral, slow development of, 
122 ; rudimentary, 283 

Origin of species, fertilisation and, 
329 ; by natural selection, 333 ; by 
response of protoplasm to environ- 
ment, 3, 50, 51, 84, seqq., 88, 103, 
seqq., 112, seqq., 116, seqq., 126, 
133, seqq. 

Ovary, arrest of, 169; growth from 
irritation of larvaj, 171, from me- 
chanical irritations, 114 ; fx'om 
pollen-tube, 170, seqq. 

Ovules, basilar, interpretation of, 74; 
of Beta (fig. 16), 73 ; emergence 
of, 195; homology of, 303, seqq.; 
foliaceous, (fig. 85) 305; meta- 
morphoses of, 305, seqq. ; of Or- 
chids, 166, seqq., 281 ; order of de- 
velopment, interpretation of, 196 ; 
origin of, 303, seqq. ; phyllody of, 
302, (fig. 85) 305, (fig. 86) 306 

Pansy, stigma and style of, (fig. 54) 

255 ; self-fertilising forms, (fig. 55) 


Pathfinders and colours, 178 

Pear, cause of obliquity at base of, 

(fig. 38) 124 ; interpretation of re- 

ceptacular tube of, 86, (fig. 22, a) 
90, (fig. 26) 94 ; effect of tension 
and weight of, upon form of, 124 
Pedicel, origin from peduncle in 

Frodium, 309 
Pelargonium, anatomy of floral recep- 
tacle, (fig. 13) 65-67 
Peloria, 128, seqq. ; causes of, 130 ; 
and generic characters, 132 ; he- 
reditary, 131 ; and hypertrophy, 
131; induced by Tingi'dac, 130 
Perianth, excrescence on, (fig. 87) 

306 ; form of, 101 
Pengynous condition, 78 
Petals, adhesion of, 78, seqq.; co- 
hesion of, 56, seqq. ; colours of, 
174, seqq. ; 253, 270, {see Colours) ; 
irritabilitv of, 158, (fig. 47) 160, 
(fig. 49) 162. See Corolla. 
Phyllody, of carpels, 302; of floral 
whorls, 301, seqq. ; of ovules, 302. 
See Ovule. 
Phyllotaxis, and aestivation, (fig. 3) 
15 ; and arrangement, 39, seqq. ; 
and number, 9, seqq. ; and origin 
of flowers, 339 
Phyteuma, cohesions of, (fig. 9) 50 
Pistil, carpels, number of, 4, 7, seqq. ; 
superposition of, 46, 47, in Cam- 
panulacece, 44 ; degeneracy in, 278 ; 
development of, rate of, 192 ; fibro- 
vascular cords of, (figs. 12, 13, 14, 
16) 64, 65, 68, 71, 73 ; metamor- 
phoses of, 295, seqq. ; movements 
of, 162 ; rationale of superposition 
in, 46, 47 ; syncarpous, 62, seqq. 
See Carpel, Gynoecium, Ovai'y, Re- 
ceptacular Tube. 
Pitcher of Nepenthes^ origin of, 146, 

Placenta, axile, 62 ; as a carpophore, 
72 ; cords of, 64-77 ; free-central, 
72, 76, (fig. 19) 77 ; hypertrophy 
of, 307 ; parietal, of Orchids, a 
sign of degeneracy, 281 
Pollen, of anemophilous flowers, 267 ; 
of cleistogamous flowers, 258, 
seqq. ; degeneracy of, 273, 276 ; of 
Orchids, 173 ; in ovules, 296 ; quan- 



tity, reduction of, 273 ; of self- 
fertilising plants, 254 

Pollen-tube, effects of, 166, 167; 
irritation due to, 164, seqq. ; in 
Orchids, 166, seqq. ; in Oxalis, 260 ; 
in Verbascum. 168 ; in Violets, 258 ; 
in Willows, 170 

Pollination and nectaries, correlation 
between, 148 

Polygamous flowers and environment, 

Pressure, effects of mechanical, 101, 
seqq., 116, seqq., 123, seqq., 156, 
seqq. ; resistance to, by cell-wall, 

Primine and secundine, foliacious, 306 

PrimulacecB, free-central placenta of, 
interpretation of, (figs. 18, 19) 76, 

Principles, general, 1 ; of variation, 4 

Protandry, cause of, 198 ; explained, 
198 ; illustrations of, 191, seqq. ; 
in Echium, (fig. 20) 82 ; and self- 
fertilisation, 272, 273, seqq. 

Protogyny, anemophily as a cause of, 
200,' 269; causes of, 199, seqq.; 
emergence and order of develop- 
ment of flowers with, 195; ex- 
plained, 198 ; inconspicuousness of 
many flowers with, 195 

Protoplasm, common to animal and 
vegetable kingdoms, and pheno- 
mena same in both, 147 ; irrita- 
bility of, to electricity, 152, to 
temperature, 153, to touch, 153, 
seqq. ; origin of species due to 
responsive powers of {see Origin of 
Species) ; transmission of effects of 
irritation by continuity of, 163 


Ranunculaceoe, arrangement in, illus- 
trations of, 21 ; symmetry in, illus- 
trations of, 21 

Receptacle, floral, anatomy of, in 
Cruciferce, (fig. 6) 32; in Helle- 
bore, (fig. 12) 64; in Ivy, (fig. 14) 

68 ; in Lysimachia, (fig. 19) 77 ; 
in Pelargonium, (fig. 13) 65 ; in 
Primula, (fig. 19) 77 

Receptacular tube, 89, seqq. ; ana- 
tomy of, in Alstroemeria, (fig. 30) 
97 ; arrested conditions in, 91, 100 ; 
with calvx foliaceous, (fig. 67) 
289 ; of Cherrv, (fig. 29) 97 ; of 
Cotoneaster, (fig. 22, 6) 90; of 
Fuchsia, (fig. 27) 94 ; of Galanthus, 
98; of Hawthorn, (fig. 25) 93; 
interpretation of, 86 ; morphologi- 
cal investigations of, 90 ; of Mus- 
scenda, (fig. 68) 290 ; of Ka7'cissus, 
98 ; of Orchids, (fig. 23) 92 ; of 
Pear, 86, (fig. 22, a) 90, (fig. 26) 
94; o£ Prunus, (fig. 28) 95; of 
Rose, (fig. 24) 93 ; teratological in- 
vestigations of, 92 ; views of, 89 

Regularity, acquired, 128 ; explained, 
5 ; observations on, 101 ; position 
of flowers with, 101 ; Tingidce as 
causing, 130. See Peloria. 

Resupination, origin of, 107 

Roots, adhesive, of Orchids, (fig. 42) 
137; origin of hairs on, 137; of 
Ivy, effects of light on, 155 

Rudimentary organs, 283 

Salvia, cleistogamous species, 262, 
263; cords of sepals of, 55; fila- 
ments of, 268; self-fertilising spe- 
cies, 261 

Scent, absence of, in self-fertilising 
flowers, 254 

Secretive tissues, as conducting, 164, 
seqq.; irritation as a cause of, 142 ; 
of milk, 147 ; as nectaries, 140, 
seqq. ; in Nepenthes, 146 ; origin 
of, 141 

Secundine, and primine, foliaceous, 

Seeds, character of, for double flowers, 
299 ; number of, compared with 
carpels, 21, 278, with stamens, 
275 ; proportion of, to seedlings in 
Orchids, 280 



Selection, constitutional, 330, 334 j 
experiment in, 331 ; by insects, 
335 ; of the luckiest, 331 ; natural, 
333. See Natural Selection. 

Self-fertilisation, and the flora of 
Dovrefjeld, 259 ; cosmopolitan, 
283 ; Mr. Darwin's views on, 215, 
and review of, 315, seqq. ; and de- 
generacy, 252 ; of Epilobium, (fig. 
53) 255; general, 192, 199, 216; 
and homomorphism, 214; illustra- 
tions of, (figs. 52-60) 255-262 ; 
injuriousness of, disproved, 315, 
seqq. ; misinterpretations regard- 
ing, 312, seqq. ; of Orchids, 253, 
318; peculiarities of, 253; rapid 
recovery of, 320; of Stellaria media, 
(fig. 52) 255 ; and whole colour- 
ing, 183 

Sensitiveness, 151. See Protoplasm. 

Sepaline cords, source of staminal 
and carpellary, 42, seqq. ; in Cam- 
panula, (fig. 8) 43, and (fig. 15) 
71 ; Labiatw increase of, in calyx 
of, 56 ; Salvia, in calyx of, 55 

Sepals, arrest of, 8 ; carpellary lobes 
of, in Pea, (fig. 70) 292 ; cords of, 
in Campanula, (fig. 8) 43, (fig. 15) 
71, and in Hollyhock, 44; de- 
velopment of, order of, 191 seqq. ; 
emergence of, 184, seqq., and in 
Crucifcrce, 32 ; foliaceous, in Ra- 
nunculus (fig. &Q\ in Tri folium 
(fig. 67), 289 : homologous with 
petioles, 288 ; lateral pair of, in 
Crucifers, first to emerge, (fig. 6) 
32, 185 ; nectaries superposed to, 
in Hellebore, (fig. 5) 22 ; numbers 
of, in whorls, 25, seqq. * ovulife- 
rous, in Violet, (fig. 71) 292 ; pe- 
taloid, one abnormallv in Linaria, 
(fig. 69) 291, normally in Mus- 
sanda, (fig. 68) 290 ; petals super- 
posed to, in Garidella, (fig. 4) 21 ; 
pistiloid, 291 ; staminoid, 291 ; ve- 
nation of, 289 

Septa, absorption of, in liber and 
wood-fibres, 250; formation of, in 
pistils, 70, seqq. 

Sex, sudden appearance of, 338 ; 
arrest of, 246 ; change of, in Calen- 
dula, 241 ; origin of, 246, 249 ; of 
seeds, 247 ; and soil, 239 ; and 
temperature, 237 

Sexuality, in Calendula, 241 ; in Cen- 
taurea, 240 ; and environment, 
230, 245 ; and heterogamy, 243 ; 
and heterostylism, 244 ; and nutri- 
tion, 233, seqq. ; and soil, 239 

Solution, explained, 5 

Spring, in corolla of Genista, (fig. 47) 
160; of stamens in Medicago, (fig. 
49) 162; of styles 125, of Viola, 
(fig. 54) 255 

Stamens, adelphous, and nectaries, 58 ; 
adhesion of, and mechanical forces, 
81 ; cohesion of, 57 ; declinate, 
110, 125, in Dictamnus, (fig. 33) 
110; in Echium, (fig. 20), 82; in 
Epilobium, (fig. 34), 111; dimorphic, 
(fig. 37) 121; distribution of forces 
in, 81, 126 ; with heterostylism, 
203, seqq. ; irregularity in, origin of, 
109 ; irritability of, 159,161 ; move- 
ment of, 162 ; metamorphoses of, 
292, 298 ; petaline, cause of absence 
of, 7, 20 ; whorls, number of, 8 

Staminode, moA'ement of, in Lopezia, 
(fig. 48) 161 

Stigmas, of anemophilous flowers, 
269; of Aristolochia, (fig. 21) 83; 
of heterostyled flowers, 216; irri- 
tability of, 115, 163; long-lived, 
269 ; movements of, 162 ; by pro- 
toplasmic continuity, 163 

Stimulus, produced by crossing, ad- 
vantages of, 330 ; temporary effect 
of, 312, 330 

Stipules, of Acacia sphcerocephala, due 
to irritation, 157 ; nectariferous, 
of Impatiens, (fig. 43) 140 

Strains, effect on boirghs,(fig, 39) 125 ; 
and cohesions, 51, 53 ; hypertrophy 
by, in pears, (fig. 38) 124, in pedi- 
cels, 123, on stems, 123, on struc- 
tures, 123 seqq. 

Struggle for existence in seedlings, 
period of greatest, 330 



Styles, hairs within, origin of, 139 ; 
of heterostyled plants, 203, seqq. ; 
movement of springs in, of Pansy, 
(fig. 54) 255 ; piston-action of, 
(fig. 11) 60; of self-fertilising 
plants, 254 

Stylopod, placental origin of, 72 

Superposition, of carpels, 44; laws 
of, 41 

Supportive tissues, 127 

Symmetry, floral, changes in, 186; 
decrease and increase of, 18; illus- 
trations of, in Banunculaceoe, 21 ; 
and phyllotaxis, 14 ; variations of, 

Syncarpous pMil, 62 

Syngenesious anthers, 59 

Tendrils, of Ampelopsis, 145 ; of 
Cucurhilacece, 145 ; thickening of, 
due to iri'itation, 156 

Teratology, 2, 285, seqq. ; 295, seqq. ; 
301, seqq. 

Teucrium, structure of flower in adap- 
tation to insects, 56, (fig. 36) 117 

Trichomes, origin of, 133, seqq. 

Trimorphic flowers, 210, seqq. 

Typical flower, diagram and structure 
of, (fig. 1) 3 


Uncaria, hook of, (fig. 46) 156 
Unions, cause of, 84 ; effect of hyper- 
trophy in, 86 ; illegitimate, 206 ; 
legitimate, 204 
Unsymmeti'ical, corolla, 5 ; deci'ease 
in floral whorls, 20 

Variation, principles of, in flowers, 4 
Vascular cords, in Campanula, (fig, 8) 

43 ; as floral units, 300, 308, 309 ; 

in Malvaceae, 43 ; origin of, 42. See 

Versatile anthers, cause of, 268 ; in 

wind-fertilised flowers, 266, seqq. 
Verticillate and opposite leaves, 91 
Vessels and cells, constructed to resist 

pressure, 127 ; as supportive, 127 
Violet, cleistogamous, (fig. 56) 258 ; 

style and stigma of, in self-fertilising 

forms, (fig. 55) 257 
Virescence, explained, 301 


"Weeds, and fertilisation, 281; self- 
fertilising, cosmopolitan, 283 

White flowers, 1 80 ; efi'ect of crossing 
with, 180 ; and self-fertilisation, 

Whorls, floral,alternation of, 39, seqq.; 
arrangement of, 39, seqq. ; examples 
of one to twelve membered, 25, 
seqq. ; illustrations from Eanun- 
culacece, 21, seqq. ; origin of, in 
CrucifercBy (fig. 6) 32 ; projected 
cycles, 38 ; superposition of, 39, 
seqq. ; symmetrical increase and 
decrease of, 18, and cause of, 19; 
of typical flower, (fig. 1) 3 

Wind-fertilised flowers. See Ane- 
mophilous and Anemophily. 

Wood-fibre, origin of, 250 

Zygoraorphism, origin of, 102, 116, 





Bain. LL. D., Professor of Logic in the University of Aberdeen. 

8vo. Cloth, ^5.00. 

The object of this treatise is to give a full and pystematic account of two 

principal divisions of the science of mind— the senses and the intellect. The 

value of the third edition ot the work is greatly enhanced by an account of the 

psychology of Aristotle, which has been contributed by Mr. Grote. 

LL. D. 8vo. Cloth, $5.00. 
The present publication is a sequel to the former one on " The Senses and the 
Intellect," and completes a systematic exposition of the human mind. 

MENTAL SCIENCE. A Compendium of Psychology and the His- 
tory of Philosophy. Designed as a Text-book for High-Schools and 
Colleges. By Alexander Bain, LL. D. 12mo. Cloth, leather 
back, $1.50. 
The present volume is an abstract of two voluminous works, "The Senses 
and the Intellect" and "The Emotions and the Will," and presents in a com- 
pressed and lucid form the views which are there more extensively elaborated. 

MORAL SCIENCE. A Compendium of Ethics. By Alexander 
Bain, LL. D. 12mo. Cloth, leather back, $1.50. 
The present dissertation falls under two divisions. The first division, en- 
titled The Theory of Ethics, gives an account of the questions or points brouiiht 
into discussion, and handles at length the two of greatest prominence, the Ethical 
Standard and the Moral Faculty. The second division— on the Ethical Systems 
—is a fall detail of all the systems, ancient and modern. 

MIND AND BODY. Theories of their Relations. By Alexander 
Bain, LL. D. 12mo. Cloth, $1.50. 
"A forcible statement of the connection between mind and body, studying 
their subtile interworkings by the light of the most recent physiological investi- 
gations." — Christian Register. 

Bain, LL. D. Revised edition. 12mo. Cloth, leather back, $2.00. 

EDUCATION AS A SCIENCE. By Alexander Bain, LL. D. 
12mo. Cloth, $1.'75. 

edition. Part I. Intellectual Elements of Style. By Alexander 
Bain, LL, D., Emeritus Professor of Logic in the University of 
Aberdeen. 12mo. Cloth, leather back, $1.50. 

ON TEACHING ENGLISH. With Detailed Examples and an 
Inquiry into the Definition of Poetry. By Alexander Bain, LL. D. 
12mo. Cloth, $1.25. 

PRACTICAL ESSAYS. By Alexander Bain, LL. D. 12mo. 

Cloth, $1.50. 

New York J D. APPLETON & CO., 1, 3, & 6 Bond Street 



THOUGHT. By Joseph Le Conte, LL. D., Professor of Geology 
and Natural History in the University of California. With numer- 
ous Illustrations. 12mo. Cloth, $1.50. 

" Much, very much has been written, especially on the nature and the evi- 
dences of evolution, but the literature is bo voluminous, much ot it 6o fragment- 
ary and most of it eo technical, that even very intelligent persons have still very 
vacTie ideas on the subject. I have attempted to give (1) a very concise account 
of what we mean by evolution, (2) an outline of the evidences of its truth drawn 
from many different sources, and ^3) its relation to fundamental leligious beliefs." 
—Extract from Preface. 

ELEMENTS OF GEOLOGY. A Text-book for Colleges and for 

the General Reader. By Joseph Le Conte, LL. D. W ith upward 

of 900 Illustrations. New and enlarged edition. 8vo. Cloth, $4.00. 

" Besides preparing a comprehensive text-book, suited to present demands, 
Professor Le Conte has given us a volume of great value as an exposition of the 
subject, thoroughly up to date. The examples and applications of the work are 
almost entirely derived from this country, so that it may be properly considered 
an American geology. We can commend this work without qualification to all 
who desire an intelligent acquaintance with geological science, as fresh, lucid, 
full, antheutic, the result of devoted study and of long experience in teaching, 
—Popular Science MontMy. 

RELIGION AND SCIENCE. A Series of Sunday Lectures on the 

Relation of Natural and Revealed Religion, or the Truths revealed 

in Nature and Scripture. By Joseph Le Conte, LL. D, 12mo, 

Cloth, $1,50. 

" We commend the book cordially to the regard of all who are interested in 
whatever pertains to the discussion of these grave questions, and especially to 
those who desire to examine closely the strong foundations on which the Chris- 
tian faith is reared."— -Bos to» Journal. 

SIGHT : An Exposition of the Principles of Monocular and Binocular 

Vision. By Joseph Le Conte, LL. D. With Illustrations. 12mo. 

Cloth, $1.50. 

" Professor Le Conte has long been known as an original investigator in this 
department : all that he gives us is treated with a master-hand. It is pleasant to 
find an American book that can rank with the very best of foreign books on 
this subject."— 2%€ Nation. 

COMPEND OF GEOLOGY. By Joseph Le Conte, LL. D. 12mo. 
Cloth, $1.40. 

New York: D. APPLETON & CO., 1, 3, & 5 Bond Street 



William Dawson, F. R. S. Vol, 61 of The International Scientific 
Series. With Illustrations. 12mo. Cloth, $]. 75. 

"The object of this work is to give, in a connected form, a summary of the 
development of the vegetable kingdom in geological time. To the geologist and 
botauist the subject is one of importance with reference to their special jjursuits, 
and one on which it has not been easy to find any couvenieut manual of informa- 
tion."— i^rom the Preface. 

TION OF ANIMALS. By Angelo Heilprin, Professor of In- 
vertebrate Paleontology at the Academy of Natural Sciences, Phila- 
delphia, etc. Vol. 57 of The International Scientific Series. One 
vol., 12mo, 435 pages, $2.00. 

" In the preparation of the following pages the author has had two objects in 
view : that of presenting to his readers such of the more significant facts con- 
nected with the past and present distribution of animal life as might lead to a 
proper conception of the relations of existing faunas ; and, secondly, that of 
furnishing to the student a work of general reference, wherein the more salient 
features of the geography and geology of animal forma cotild be sought after 
and readily found."— i^'rom the Preface. 

ANIMAL MAGNETISM. From the French of Alfred Binet and 

Charles Fere. Vol. 59 of The International Scientific Series. 12mo. 

Cloth, $1.50. 

"The authors, after giving a brief, clear, and instructive history of animal 
magnetism from its remotest known origin down through Mesmer and the Aca- 
demic period to the present day, record their personal investigations among the 
hysterical, nervous, and generally supersensitive female patients in the great 
Paris hospital. La Salpetriere, of which M. Fere is the assistant physician."— 
Jourrial of Commerce. 



Abercromby, Fellow of the Royal Meteorological Society, London. 

Vol. 58 of The International Scientific Series. 12mo. Cloth, $1.75. 

" "Mr. Abercromby has for some years made the weather of Great Britain a 
special study, and has recently extended his experience by making a meteorologi- 
cal tour around the world. Ab a fruit of this preparation, he gives us a book that 
is to be commended for its simple, deliberate style, freedom from technicality and 
unnecessary theorizing, rational description, classification, and explanation of 
atmospheric phenomena, and rich store of iUustration from the weather-maps of 
many parts of the world."— 2%€ Nation. 

New York: D. APPLETON & CO., 1, 3, & 5 Bond Street. 




Revised edition, with Additions. 12mo. Cloth, $2,00. 


TO SEX. With many Illustrations. A new edition. 12mo. 
Cloth, $3.00. 

ROUND THE WORLD. New edition. 12mo. Cloth, $2.00. 

LOWER ANIMALS. 12mo. Cloth, $3.50. 

DOMESTICATION. With a Preface, by Professor Asa Gray. 
2 vols. Illustrated. Cloth, $5.00. 

INSECTIVOROUS PLANTS. 12mo. Cloth, $2.00. 


With Illustrations. 12mo. Cloth, $1.25. 

ARE FERTILIZED BY INSECTS. Revised edition, with 
Illustrations. 12mo. Cloth, $1.75. 



THE SAME SPECIES. With Illustrations. 12mo. Cloth, 

Darvtin, LL. D., F. R. S., assisted by Francis Darwin. With Illus- 
trations. 12mo. Cloth, $2.00. 

THE ACTION OF WORMS. With Observations on their 
Habits. With Illustrations. 12ino. Cloth, $1.50. 

New York: D. APPLETON & CO., 1, 3, & 5 Bond Street. 



Cloth, $1.50. 

OF ZOOLOGY. With 82 Illustrations. 12mo. Cloth, $1.75. 

45 cents. 

MAN'S PLACE IN NATURE. 12mo. Cloth, $1.25. 

ON THE ORIGIN OF SPECIES. 12mo. Cloth, $1.00. 

NIHILISM. 12mo. Limp cloth, 50 cents. 

ANIMALS. Illustrated. 12mo. Cloth, $2.60. 

ANIMALS. 12mo. Cloth, $2.50. 


Cloth, $1.75. 

CRITIQUES AND ADDRESSES. 12mo. Cloth, $1.50. 

STUDY OF BIOLOGY. 12mo. Cloth, $1.25. 

NATURE. With Illustrations and Colored Plates. 12mo. Cloth, 

OGY AND HYGIENE. By T. H. Huxley and W. J. Youmans. 

12mo. Cloth, $1.50. 

New York: D. APPLETON & CO., 1, 3, & 6 Bond Street. 




in Relation to Putrefaction and Infection. 12mo. Cloth, §1.50. 

ON FORMS OF WATER, in Clouds, Rivers, Ice, and Glaciers 
With 35 Illustrations. 12mo. Cloth, $1.50. 

HEAT AS A MODE OF MOTION. New edition. 12mo. 

Cloth, $2.50. 

ON SOUND: A Course of Eight Lectures delivered at the Royal 
Institution of Great Britain. Illustrated. 12mo. New edition. 
Cloth, $2.00. 

PIiE. 12mo. New revised and enlarged edition. Cloth, $2.50. 

LIGHT AND ELECTRICITY. 12mo. Cloth, $1.25. 

LESSONS IN ELECTRICITY, 1875-'76. 12mo. Cloth, $1.00 

12mo. Cloth, $2.00. 

FARADAY AS A DISCOVERER. A Memoir. 12mo. Goth, 


main of Radiant Heat. $5.00. 

SIX LECTURES ON LIGHT. Delivered in America in 18*72- 
'73. With an Appendix and numerous Illustrations. Cloth, $1.50. 

FAREWELL BANQUET, given to Professor Tyndall, at DeU 
monico's, New York, February 4, 1873. Paper, 50 cents. 

ADDRESS delivered before the British Association, assembled at Bel- 
fast. Revised with Additions, by the author, since the DeUvery 
12mo. Paper, 50 cents. 

New York : D. APPLETON & CO., 1, 3, & 5 Bond Street 



a Research on Primitive Nervous Systems. 12mo. Cloth, fl.VS. 

" A-lthoagh I have throughout kept in view the requirementa of a general 
reader, I have also sought to render the book of service to the working physi- 
ologist, by bringing together in one consecutive account all the more important 
observations and results which have been yielded by this research."— ^iciraci 
from Preface. 

"A profound research into the laws of primitive nervous systems conducted 
by one of the ablest English investigators. Mr. Komanes set up a tent on the 
beach and examined his beantitul pets tor six summers in succession. Such 
patient and loving work has borne its fruits in a monograph which leaves 
nothing to be said about jelly-fish, star-fish, and sea-urchins. Every one who 
has studied the lowest forms of life on the sea-shore admires these objects. But 
few have any idea of the exquisite delicacy of their structure and their nice 
adaptation to their place in nature. Mr. Komanes brings out the subtile beauties 
of the rudimentary organisms, and shows the resemblances they bear to the 
higher types of creation. His explanations are made more clear by a large 
number of illustrations. While the book is well adapted for popular reading it 
is of special value to working physiologists."— iVew York Journal of Commerce. 

"A most admirable treatise on primitive nervous systems. The subject-matter 
is full of original investigations and experiments upon the animals mentioned as 
types of the lowest nervous developments."— 5osto« Commercial Bulletin. 

" Mr. George J. Romanes has already established a reputation as an exact and 
comprehensive naturalist, which his later work, 'Jelly -Fish, Star-Fish, and Sea- 
Urchius,' fully confirms. These marine animals are well known upon our coasts, 
and always interest the on-Iookers. In this volume (one of the ' International 
Scientific Series ') we have the whole story of their formation, existence, nervous 
system, etc., made most interesting by the simple and non-professional manner 
of treating the subject. Illustrations aid the text, and the professional student, 
the naturalist, all lovers of the rocks, woods, and shore, as well as the general 
reader, will find instruction as well as delight in the narrative."— £o«to;i Com- 

ANIMAL INTELLIGENCE. 12mo. Cloth, $1.75. 

"A collection of facts which, though it may merely amuse the unscientific 
reader, will be a real boon to the student of comparative psychology, for this is 
the first attempt to present systematically the well-assured results of observation 
on the mental life of animals."— iSo^wr-t^ay Review. 

Essay on Instinct, by Charles Darwin. 12mo. Cloth, $2.00. 

"Mr. Romanes has followed up his careful enumeration of the facts of ' Ani- 
mal Intelligence,' contributed to the ' International Scientific Series,' with a 
work dealing with the successive stages at which the various mental phenomena 
appear in the scale of life. The present installment displays the same evidence 
ot industry in collecting facts and caution in co-ordinating them by theory as the 
lormer.' '— The Athenaeum. 

New York : D. APPLETON & CO., 1, 3, & 5 Bond Street. 


BODY AND WILL: Being an Essay concerning Will in its 
Metaphysical, Physiological, and Pathological Aspects. 12mo. 
Cloth, $2.50. 

BODY AND MIND : An Inquiry into their Connection and Mutual 
Influence, specially in reference to Mental Disorders. 1 vol., 12mo. 
Cloth, $1.50. 


PHYSIOLOGY OF THE MIND. New edition. 1 vol., 12mo. 
Cloth, $2.00. Contents : Chapter L On the Method of the Study 
of the Mind. — II. The Mind and the Nervous System. — III. The 
Spinal Cord, or Tertiary Nervous Centres ; or. Nervous Centres 
of Reflex Action. — IV. Secondary Nervous Centres ; or, Sensory 
Ganglia ; Sensorium Commune. — V. Hemispherical Ganglia ; Cortical 
Cells of the Cerebral Hemispheres; Ideational Nervous Centres; 
Primary Nervous Centres ; Intellectorium Commune. — VI. The 
Emotions. — VII. Volition. — VIII. Motor Nervous Centres, or Mo- 
torium Cummune and Actuation or Effection. — IX. Memory and 

PATHOLOGY OF THE MIND. Being the Third Edition of 
the Second Part of the " Physiology and Pathology of Mind," 
recast, enlarged, and rewritten. 1 vol., 12mo. Cloth, $2.00. Con- 
tents : Chapter I. Sleep and Dreaming. — II. Hypnotism, Somnam- 
bulism, and Allied States. — III. The Causation and Prevention of 
Insanity: (A) Etiological. — IV. The same continued. — V. The 
Causation and Prevention of Insanity : (B) Pathological. — VI. The 
Insanity of Early Life. — VII. The Symptomatology of Insanity. — ■ 
VIII, The same continued. — IX. Clinical Groups of Mental Disease. 
— X. The Morbid Anatomy of Mental Derangement. — XI. The 
Treatment of Mental Disorders. 

Scientific Series.) 1 vol., 12mo. Cloth, $1.50. 

" The author is at home iu his Bubject, and presents hie views in an almost 
singularly clear and satis^factory manner. . . . The volume is a valuable contri- 
bution to one of the most difficult and at the same time one of the most impor- 
tant subjects of investigation at the present day."— iV(?w York Observer. 

" Handles the important topic with masterly power, and its suggestions are 
practical and of great sraXue.''''— Providence Press. 

New York : D. APPLETON & CO., 1, 3, & 5 Bond Street 



CONDITION OF SAVAGES, Fourth edition, with numerous Ad- 
ditions, With Illustrations, 8vo. Cloth, |5.00. 

"The first edition of this work was published In the year 1870. The work 
hag been twice revised for the press in the interval, and now appears in its 
fourth edition enlarged to the extent of nearly two hundred vages, including a full 

"This interesting work— for it ia intensely so in its aim, scope, and the abil 
ity of its author— treats of what the scientists denominate anthropology, or tho 
natural history of the human species ; the complete science of man, body and 
soul, including sex, temperament, race, civilization, eiG.'''— Providence Press. 

SAVAGES. Illustrated. Entirely new revised edition. 8vo. Cloth, 

The book ranks among the noblest works of the interesting and Important 
class to which it belongs. As a resume of our present knowledge of prehistoric 
man, it leaves nothing to be desired. It is not only a good book of reference but 
the best on the subject. 

" This is, perhaps, the best summary of evidence now in our possession con- 
cerning the general character of prehistoric times. The Bronze Age, The Stone 
Age, The Tumuli, The Lake Inhabitants of Switzerland, The Shell Mounds, Tho 
Cave Man, and The Antiquity of Man, are the titles of the most important chap- 
ters."— Z>r. (J. K. Adams's Manual of Historical Literature. 

ANTS, BEES, AND WASPS. A Record of Observations on the 
Habits of the Social Hymenoptera. With Colored Plates. 12mo. 
Cloth, $2.00. 

"This volume contains the record of various experiments made with ants, 
bees, and wa?ps during the last ten years, with a view to test their mental con- 
dition and powers of sense. The principal point in which Sir John's mode of 
experiment differs from those of Huber, Forel, McCook, and others, is that he 
has carefully watched and marked particular insects, and has had their nests 
under observation for long periods— one of his ants' nests having been under 
constant inspection ever since 1874. His observations are made principally upon 
ants, because they show more power and flexibility of mind ; and the value of 
his studies is that they belong to the department of original research." 

" We have no hesitation in saying that the author has presented us with the 
most valuable series of observations on a special subject that has ever been pro- 
duced, charmingly written, full of logical deductions, and, when we consider his 
multitudinous engagements, a remarkable illustration of economy of time. A* a 
contribution to insect psychology, it will be long before this book finds a Bar* 
allel." — London Athenaeum. 

New York : D. APPLETON & CO., 1, 3, & 5 Bond Street. 


engravings, diagrams and maps in color, and including a separate 
chapter on the geological history and the physical features of the 
United States. By John D. Quackenbos, A. M., M. D., Adjunct 
Professor of the English Language and Literature, Columbia College, 
New York, Literary Editor; John S. Newberry, M. D., LL. D., 
Professor of Geology and Paleontology, Columbia College ; Charles 
H. Hitchcock, Ph. D., Professor of Geology and Mineralogy, Dart- 
mouth College ; W. Le Conte Stevens, Ph. D., Professor of Physics, 
Packer Collegiate Institute ; Henry Gannett, E. M., Chief Geog- 
rapher of the United States Geological Survey ; William H. Dall, 
of the United States National Museum; C. H*rt Merriam, M. D., 
Ornithologist of the Department of Agriculture , Nathaniel L. Brit- 
ton, E. M., Ph. D., Lecturer in Botany, Columbia College ; George 
F. KcNz, Gem Expert and Mineralogist with Messrs. Tiffany & Co., 
New York; Lieutenant George M. Stoney, Naval Department, 
Washington. Large 4to. Cloth, $1.90. 


Consisting of General Maps of the United States and Territories, 
and a County Map of each of the States, all printed in Colors, 
together with Railway Maps and Descriptive Text Outlining the 
History, Geography, and Political and Educational Organization of 
the States, with latest Statistics of their Resources and Industries. 
Imperial 8vo, cloth. $1.50. 

Reclus. Translated and edited by E. G. Ravenstein. With nu- 
merous Illustrations, Maps, and Charts. 
M. Reclus the distinguished French Geographer has given in this work 

the most thorough and comprehensive treatise on the countries of the 

world yet produced. Maps, plans, and illustrations are lavish. It is 

subdivided as follows : 

Europe, in 5 volumes. Imperial 8vo. 
Asia, in 4 volumes. Imperial 8vo. 
Africa, in 3 volumes. Imperial Svo. 
America. {In preparation.) 

Price, $6.00 per volume in library binding. Sold only by subscrip- 

In two volumes. Vol. I. The Earth. Vol. II. The Ocean, Atmos- 
phere, and Life. With Maps and Illustrations. Price, $6.00 per 
volume, library binding. Sold only by subscription. 

New York: D. APPLETON & CO., 1, 3, & 5 Bond Street 

New York Botanical Ga''."^®" ,,Vmi^»1[ 

• '^*liH 


THK ^ 




.:»vX. ■.-,■. ■ 



t ilL 1 ^^^^^^^1