> 
 
 * s 
 
 
 
 &W 
 
 ■4* 
 
 : 1. 
 
 
 
 *& 
 
 
 
 & 
 &* 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 4/ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 A 
 
 
 
 
 
 
 
 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 .1 "> 
 
Galls dLivd Insects Producing Them 
 
 By 
 
 Melville Thurston Cook, A. M. 
 
 r ■ ■ '-* • • 9 
 
 Presented to the Faculty of the College of Arts, Philosophy 
 and Science, Ohio State University, as the thesis requirement 
 for the degree of Doctor of Philosophy. June 1904. 
 
• Cs 
 
 MAR 
 
Contents 
 
 Part I. The Morphology of Leaf Galls. 
 
 Part II. Apical Bud Galls. 
 
 Part III. Lateral Bud Galls. 
 
 Part IV. Stem Galls. 
 
 Part V. Development of Galls. 
 
 Part VI. Flower and Fruit Galls. 
 
 Part VII. Root Galls. 
 
 Part VIII. Histology of Galls. 
 
 Part IX. The Ovipositors and Mouth Parts of Gall Producing 
 
 Insects. 
 Appendix. 
 
 P^K^ 
 
 Parts I and II published in Ohio Naturalist Vol. II. pp 
 263-278. 1902 
 
 Parts III. IV. V. published in Ohio Naturalist Vol. 
 III. pp 419-436. 1903. 
 
 Parts VI. VII. VIII. IX and appendix published in 
 Ohio Naturalist Vol. IV pp 115-147. 1904. 
 
[Rkprinted from Ohio Naturalist, Vol. II, pp. 263-27S, 1902.] 
 
 GALLS AND INSECTS PRODUCING THEM. 
 
 Melville Thurston Cook. 
 Part i. The Morphology of Leap Gales. 
 
 The purpose of this study was to contribute to the knowledge 
 of cellular activity of the plant under peculiar animal stimulus ; 
 to compare the effects of the two sets of insect organs, mouth 
 parts and ovipositors, and to throw additional light on the classi- 
 fication. The statements made in this paper are based on a large 
 number of collections. The collection of stem galls was too 
 incomplete to draw conclusions and is therefore reserved for a 
 future paper. No attempt was made to follow the development 
 of the galls but rather to make a comparison of the structure of 
 the various forms of galls. 
 
 My paper was practically complete before I received the papers 
 of H. Fockeu. After receiving his paper I reviewed my own to 
 determine wherein my results agreed with or varied from his 
 conclusions. Experiments such as are described by H. Fockeu 
 to ascertain the cause of the gall formation were not attempted. 
 
 Fockeu' s studies were grouped according to the plants affected ; 
 my own studies were grouped with reference to the insect pro- 
 ducing the galls. 
 
 METHODS. 
 
 For the killing and fixing, several fluids were used, but the 
 most successful were Chromo-acetic and Picric-alcohol. A num- 
 ber of different stains were used, but Delafields-Haemotoxylon 
 proved very satisfactory for most work. 
 
 For the drawings a Bausch & Lomb microscope and camera 
 lucida were used ; for the normal leaf, a i-inch ocular and a 
 ^-inch objective, and for the galls a i-inch ocular and a ^3-inch 
 objective. Since it was unnecessary to make drawings of the 
 entire galls, drawings were made from one or more parts to show 
 the characteristic structure, and this part is indicated on the small 
 diagrammatic drawings. Since the galls were so variable in size, 
 it was practically impossible to make the diagrammatic drawings 
 on a definite scale. 
 
 GENERAL CLASSIFICATION. 
 
 As a matter of convenience the following temporary classifica- 
 tion, based on location of the galls was adopted for this and other 
 
264 The Ohio Naturalist. [Vol. II, No. 7, 
 
 papers now in preparation: A. Stem galls; B. Leaf galls; 
 C. Bud galls, a. Terminal buds, b. Lateral buds ; D. Root galls. 
 
 Leaf galls may in many cases be classed as bud galls if we con- 
 sider that the egg in some orders of insects is deposited while the 
 leaf is in the bud, but in the above classification the term applies 
 to the developed gall, and the ' bud gall ' applies to a distortion 
 of the entire bud. 
 
 i. The Normal Leaf Structure and Its Variations. 
 The normal leaf structure may be said to consist of a single layer 
 of epidermis on the upper and lower surfaces of the leaf ; next to 
 the upper epidermis is the usually single layer of palisade or 
 columnar cells, placed with their long axis at right angles to the 
 surface of the leaf ; between the palisade cells and the lower epi- 
 dermis is the mesophyll, made up of many layers of irregular 
 cells, between which are the large air spaces connected with the 
 outside by the stomata in the lower epidermis ; running through 
 the leaf are the fibro-vascular bundles noticable to the naked eye 
 as the venation. 
 
 Although the above may be said to be a description of a typical 
 leaf, it must be kept in mind that leaves are subject to great 
 variation and this must be taken into consideration in a discus- 
 sion of the variation of the gall structure from the normal leaf. 
 The structure of the gall must be compared with the structure 
 of the normal leaf of the plant on which the gall is found, not 
 with the typical leaf. 
 
 A brief study of the normal leaves of the plant will serve to 
 emphasize the preceding points. Hicoria ovata ( Mill. ) Britton 
 (Fig. i), Uhiuis americana L. (Fig. 4), and Tilia americana L. 
 (Fig. 6) may be considered as typical and yet in themselves show 
 minor differences. In Vitis vulpina L. (Fig. 3) the palisade is 
 not so pronounced as in the preceding and the mesophyll is more 
 compact. In Quercus alba L. (Fig. 7) and in Acer saccharin um 
 L. (Fig. 5) the palisade is typical, but the mesophyll is very 
 compact. In Sah'x cordata Muhl. (Fig. 2) the mesophyll while 
 distinct from the palisade has assumed palisade characters. 
 
 The differences in structure between the normal leaves of 
 Hicoria ovata (Fig. 1) and Salix cordata (Fig. 2), members of two 
 related families, are as great as those differences frequently found 
 between a normal leaf and the galls occurring upon it, e. g., H. 
 ovata (Fig. 1) and the simpler Phylloxera galls (Figs. 16-20). 
 
 2. Phytoptus Galls. This discussion is based not only on 
 the four galls described below, but from observations of several 
 others. However, the following will illustrate all the points 
 observed : 
 
 The Phytoptus galls are small and may extend on either or 
 both sides of the leaf. The outer surface of the galls show the 
 normal epidermis and below this cells which are not palisade but 
 
May, 1902.] Galls and Insects Producing Them. 265 
 
 which are elongated with the surface of the gall, i. <?., the direc- 
 tion of growth (Figs. 8, 9, 11). Projecting into the gall cavity 
 are masses of irregular shaped cells (Figs. 8-1 1). In young galls 
 these cells show a nucleus, take the stain readily and show 
 indications of maturity (Figs. 9, 11). Trichomes are always 
 found extending from the walls of the cavity (Figs. 8-1 1) of 
 young galls, but disappear as the galls approach maturity. In 
 these galls we evidently have a repeated puncturing of cells by 
 the animal and an increased activity on the part of the plant in 
 its effort to recover from the wound, the wound never being suf- 
 ficient to cause the death of that part of the plant. 
 
 My results on the Phytoptus galls agree with those of H. 
 Fockeu, except in minor points. 
 
 3. The Aphididae Gaels. In this family we find the 
 simplest form of galls discussed in this paper, of which Schizonc- 
 ura americana Riley (Fig. 12) may be taken as a type. In fact 
 it is a mere curling of the leaf and not what is usually considered 
 a gall. According to E. Perris it would be classed as a galloide. 
 However, the structure is very similar to that of a typical gall of 
 this family of insects and I see no reason why it should not be 
 considered a true gall. 
 
 When compared with the normal leaf of U. americana L. 
 (Fig. 4) the palisade cells are observed to have lost their identity 1 
 and to have assumed mesophyll characters and the mesophyll has | 
 become more compact, both distortions being characteristic of 
 true galls of this family ( Figs. 13-21 ). 
 
 In Coloplia ulmicola Fitch (Fig. 13 a. b. ) and Pemphigus ulmi- 
 fusus (Walsh.) Oestlund (Fig. 14 a. b.) both of which are also 
 characteristic galls on the elm, we find practically the same 
 structure as in S. americana. In both the outer (upper) epi- 
 dermis is much elongated ; the same being true of the inner 
 (lower) epidermis of C. ulmicola, but not in P. ulmi-fusus. The 
 identity of the palisade cells is entirely lost, the cells now being 
 slightly elongated parallel to the surface of the gall. The 
 mesophyll cells are more compact than in S. americana and far 
 more compact than in a normal leaf (Fig. 4). 
 
 A granular, dark brown, often black substance in the cells was 
 characteristic of the elm and other galls of this group. This was 
 probably tannin, and its presence seemed to depend on the host 
 plant rather than on an insect producing the gall. 
 
 The Hormaphis hamamelis Fitch ( Fig. 15 a. b. ) on the 
 Hamamelis virginiana L,. showed the same general structure as 
 the preceding galls of this order, except that the epidermal cells 
 were not so much elongated and in the inner (lower) epidermis 
 the cells were much smaller and showed thicker walls, and the 
 dark granular contents of certain cells was restricted to layers 
 near the outer (upper) surface. 
 
266 The Ohio Naturalist. [Vol. II, No. 7 r 
 
 The Phylloxera galls show considerable variation from each 
 other. P. c. avenae Fitch, P. c. fallax Riley, and P. c. globuli 
 Walsh. (Figs. 16-18), of Hicoria ovata may be taken as forming 
 a rather well defined group and as showing greatest resemblance 
 to the preceding galls of this family. When compared with the 
 normal leaf (Fig. 1) of the host, H. ovata, they show a reduc- 
 tion in size of the epidermal cells, the palisade cells losing their 
 identity, and the mesophyll becoming very compact. Very little 
 of the dark cell contents characteristic of the preceding galls of 
 this family was present, the greatest amount being formed in 
 P. c. avenae (Fig. 16 ) where it is restricted to the epidermis and 
 to the cells just below it. The cells are even less elongated and 
 more irregular than in the preceding galls. In general it may be 
 said that in this group the largest cells are midway between the 
 two layers of the epidermis and gradually decrease as we approach 
 the surfaces. This is especially true of P. c. globuli (Fig. 18 ). 
 
 P. c. spinosa Shinier (Fig. 19 a. b. ) is a very large gall occur- 
 ring on leaf, petiole, or young, green twigs of Hicoria ovata and 
 shows considerable variation from the preceding. Two zones are 
 very distinct ; the outer is composed of large cells which do not 
 take the stain readily, the inner zone of small cells stained very 
 readity and show great activity. This may, however, have been 
 due to the fact that my specimens of this gall were much younger 
 than of the preceding Phylloxera galls. A long tube for the exit 
 of the insect is formed. 
 
 In P. c. depressa Shinier (Fig. 20 a. b. ) of H. ovata and P. 
 vastatrix Planchon (Fig. 21 a. b. ) of Vitis vulpina we have still 
 other and more marked variation. The cavity is much. smaller, 
 the walls much thicker than in the preceding, and a long tube, 
 especially in P. c. depressa is formed for the exit of the insect. 
 In both cases the size of the epidermal cells is much reduced when, 
 compared with the normal (Fig. 1, 3), the palisade cells have not 
 so completely lost their identity as in the preceding and there 
 appears to be a general elongation of the cells with their long 
 axis perpendicular and not parallel to the surface of the gall. 
 A small but definite, deeply staining zone of cells surrounds the 
 cavity in P. c. depressa. Many cells show dark contents similar 
 to that found in the galls on Ulmus and Hamamelis (Fig. 12-15). 
 
 P. vastatrix shows a comparatively large number of trichomes, 
 especially near the opening, but this is probably a characteristic 
 of the host plant rather than of the gall. 
 
 The presence of the two well defined zones, which may be con- 
 sidered protective and nutritive in P. c. spinosa and P. c. de- 
 pressa, show a very marked resemblance to the Cynipidae galls 
 (Figs. 25-30). 
 
 It may be that all young galls show this arrangement into two- 
 or three zones. 
 
May, 1902.] Galls and Insects Producing Them. 267 
 
 In P. c. depressa (Fig. 20) and in P. vastatrix (Fig. 21) the 
 small larval chamber and general arrangement of the cells is very 
 similar to the leaf galls produced by Cecidomyia verrucola (Fig. 2.) 
 
 4. The Cecidomyia Gales. This group of galls shows con- 
 siderable variation. C. gleditsiae O. S. (Fig. 22 a. b. c. d. ) of 
 Gleditschia triacanthos may be taken as a type of one of the 
 simplest. In this the margins of the leaflets are in contact so as 
 to form a more or less sperical body. To the naked eye it pre- 
 sents no other distortion. Under the microscope the cells show 
 an elongation from midrib to margin, i. e., parallel to the surface 
 of the gall except near the margin, where they are irregular. 
 
 C. quercus-pilulae Walsh. (Fig. 23 a. b. ) shows a more highly 
 developed gall structure. The epidermal layers are made up of 
 smaller cells than the normal leaf. The mesophyll has lost its 
 identity and assumed the palisade structure, the long axis being 
 perpendicular to the surface of the gall. The larval chamber is 
 large and rather irregular and indefinite, and resembles a large 
 inter-cellular space. 
 
 C. verrucola O. S. (Fig. 24 a. b. ) on Tilia americana shows a 
 much higher complexity than either of the preceding. The epi- 
 dermis is made up of small cubical cells. The differentiation into 
 palisade and mesophyll is entirely lost, the cells are very irregu- 
 lar, but show a tendency to elongation at right angles to the 
 surface of the gall. The larval chamber is small and well defined. 
 
 C. q.-pilulae (Fig. 23) and C. verrucola (Fig. 24), especially 
 the latter show a striking resemblance to the more highly devel- 
 oped Phylloxera galls such as P. c. -depressa (Fig. 20) and P. 
 vastatrix (Fig. 21). 
 
 5. The Cynipidae Galls. This family presents the most 
 striking series of evolutionary development of any family studied 
 and is also apparently the most highly developed. 
 
 The general characters presented by these galls are small, 
 cubical epidermal cells ; loss of differentiation between palisade 
 and mesophyll cells, all having assumed an irregular character ; 
 a differentiation into two well defined zones of cells, the outer 
 made up of large, non-staining cells, the inner made up of smaller, 
 deeply staining cells and surrounding the larval chamber. 
 
 Fockeu divides these into four zones, which he designates as 
 follows: 1. Epidermis; 2. Parenchyma; 3. Protective; 4. 
 Nutritive ("Masse alimentaire "). These four zones may be 
 easily traced in most of our American forms, but in some they 
 show very indistinctly. 
 
 Neuroterus irregularis O. S. (Fig. 25 a. b. ) is a small, fleshy, 
 solid, irregular gall projecting from both sides of the leaf. It is 
 covered with dense growth of trichomes and contains several 
 larval chambers. In structure it does not correspond to the pre- 
 ceding description, as well as the galls described in the latter part 
 
268 The Ohio Naturalist. [Vol. II, No. 7, 
 
 of this paper. The parenchyma is divided into two very distinct 
 zones, the larval chamber occupying the lower part of the inner 
 zone. The inner zone cells have much thinner walls than those 
 of the outer cells. Surrounding the larval chamber is a zone of 
 cells which stain very deeply and probably furnish nourishment 
 to the larva. The epidermal cells are small. 
 
 Callirhytis tumifica O. S. (Fig. 26 a. b.) is a small, fleshy, solid 
 gall projecting on both sides of the leaf and resembles N. irregu- 
 laris (Fig. 25), except that it is a little larger, does not have so 
 many larval chambers and is smooth. It presents the simplest 
 characters studied, showing the characteristic small, more or less 
 cubical epithelial cells, the lack of differentiation into palisade 
 and mesophyll, and the two zones. The outer zone is very thick 
 and is in contact with the inner zone. The inner zone is narrow 
 and lies near the large larval chamber. At the point of union of 
 the two zones the cells are very small. The outer zone can be 
 readily subdivided into epidermis and parenchyma, but the inner 
 zone cannot be subdivided into two sub-zones unless we consider 
 the layer of small cells as the protective sub-zone. However, 
 this sub-zone of small cells does not possess the sclerenchyma 
 character described by Fockeu for the Cynipidae galls. 
 
 Holcaspis centricola O. S. (Fig. 27 a. b. c. ) is a large, spherical 
 gall projecting both above and below the leaf. In this we have 
 the two zones, but each retaining the characters previously 
 described ; the cells of the inner zone, however, being smaller 
 than in C. tumifica. The epidermal cells have thicker walls than 
 in any other Cynipidae gall examined. The two zones are con- 
 nected by fibro-vascular bundles. In this the four zones of 
 Fockeu are quite well defined : The outer zone forming the very 
 distinct epidermis and parenchyma ; the inner zone showing a 
 fairly well defined protective and nutritive part. 
 
 Amphibolips inanis O. S. (Fig. 28 a. b.) shows a very striking 
 resemblance to H. centricola (Fig. 27), except that it is much 
 larger. The epidermal cells do not have such thick walls as in 
 H. centricola and are much longer and narrower. The inner zone 
 is readily subdivided into the protective and nutritive sub-zones 
 described by Fockeu. The inner or nutritive sub-zone is made 
 up of thin-walled cells with prominent nuclei, the outer or pro- 
 tective sub-zone of sclerenchyma cells. The connection between 
 the two main zones is by means of fibro-vascular bundles, the 
 same as in H. centricola. 
 
 Dryophanta pahistris O. S. (Fig. 29 a. b. c.) presents a condi- 
 tion very similar to the two preceding galls, H. centricola (Fig. 
 27) and A. inanis (Fig. 28), except that the fibro-vascular bundle 
 connection between the two zones is not present ; the inner zone 
 containing the larva forms a sphere which is free in the large 
 chamber formed by the outer zone. 
 
May, 1902.] Galls and Insects Producing Them. 269 
 
 The inner zone shows a marked resemblance to H. eentricola 
 (Fig. 27). The subdivision into protective and nutritive parts in 
 my specimens was not like the characteristic zones described by 
 Fockeu ; the inner cells were apparently much thicker walled and 
 more indefinite. However, I believe that younger galls would 
 have shown the typical characters. The outer zone is thicker 
 than in either H. eentricola (Fig. 27 ) or A. inanis (Fig. 28), but 
 not so thick as in C. tumifica (Fig. 26). It can be readily sub- 
 divided into epidermis and parenchyma and it also shows a fairly 
 well defined endodermis, and in that respect differs from either 
 H. eentricola or A. inanis. 
 
 Callirhytis papillatus O. S. (Fig. 30 a. b. c), which is similar 
 to the preceding Cynipidae galls, but shows considerable varia- 
 tion from them. It is smaller than any of the preceding and is 
 embedded in the leaf very similar to C. tumifica (Fig. 26). The 
 two zones are separated, the outer being similar to A. inanis 
 (Fig. 28), the inner zone surrounding two or three larval cham- 
 bers instead of one. Next to the larva the cells are very large 
 and thin and may be considered nutritive ; outside these we have 
 well defined parenchyma or protective cells, and outside these we 
 have two or three layers of cells well filled with protoplasm. The 
 connection between the outer and inner zones is by single elon- 
 gated cells, which are very rich in protoplasm. 
 
 The evolutionary development of the preceding Cynipidae galls 
 is evident. All show the two well defined zones, the outer non- 
 staining made up of epidermis and parenchyma and the inner 
 which takes the stain readily and is made up of two subdivisions, 
 protective (or sclerenchyma cells) and nutritive (or parenchyma 
 cells). In C. tumifica (Fig. 26) we have the two zones in con- 
 tact ; in H. eentricola (Fig. 27) and in A. inanis (Fig. 28) we 
 have a separation of the two zones which are now connected by 
 fibro- vascular bundles ; in C. papillatus (Fig. 30) the two zones 
 are connected by long, undivided cells ; in D. palustris (Fig. 29 ) 
 we have a complete separation of the two zones. 
 
 With the exception of N. irregularis (Fig. 25) and C. tumifica 
 ( Fig. 26) they all show a division into four zones as described by 
 Fockeu. However, Fockeu does not describe a separation 
 between the parenchyma and protective zones which is so charac- 
 teristic of some of our American galls. I am inclined to consider 
 our American Cynipidae galls as having reached a higher stage 
 of development than the European forms. 
 
 The larva in all species evidently draws its nourishment directly 
 from the inner zone. In H. eentricola (Fig. 27) and A. inanis 
 ( Fig. 28 ) the inner zone evidently gets its nourishment through 
 the fibro- vascular bundles ; in C. papillatus (Fig. 30) the supply 
 of nourishment comes through the long filamentous cells ; in 
 D. palustris (Fig. 29) it is probable that the larva is far advanced 
 
270 The Ohio Naturalist. [Vol. II, No. 7, 
 
 in its development before the separation of the two zones and the 
 nourishment remaining in the inner zone at the time of the separ- 
 ation is sufficient to complete its development. 
 
 Adler and Stratton after describing similar modifications in the 
 European Cynipidae galls, say: "Besides these histological 
 differences, the outward characters are also of varying complex- 
 ity ; each infinitesimal improvement, which has been of service 
 as a protection against parasites, or has been successful in secur- 
 ing natural conditions favorable to the life and growth of the 
 larva, has been preserved, and has formed the starting point of 
 further beneficial variation. It is always that larva which has 
 been able to induce successful morphological abnormalition, which 
 is reproduced to continue the race ; the unsuccessful perish. The 
 ruling force is natural selection ; it is impossible that intelligence 
 or memory can be of any use in guiding the Cynipidae ; no 
 Cynips ever sees its young, and none ever pricks a bud the sec- 
 ond season, or lives to know the results that follow the act. 
 Natural selection alone has preserved an impulse which is released 
 by seasonally recurring feelings, sights, or smells,* and by the 
 simultaneus ripening of the eggs within the fly. These set the 
 whole physiological apparatus in motion, and secure the insertion 
 of eggs at the right time and in the right place. The number of 
 eggs is instinctively proportionate to the space suitable for ovipo- 
 sition, to the size of the fully grown galls, and to the food sup- 
 plies available for their nutrition." 
 
 conclusions. 
 
 1. Galls may be classified into two general groups, viz., those 
 produced by mouth parts and those produced by oviposition. 
 Those produced by oviposition may be considered the more highly 
 developed. 
 
 2. The family Cynipidae shows by far the highest develop- 
 ment of gall structures. 
 
 3. The morphological character of the gall depends upon the 
 genus of the insect producing it rather than upon the plant on 
 which it is produced ; i. e. , galls produced by insects of a partic- 
 ular genus show great similarity of structure even though on 
 plants widely separated ; while galls on a particular genus of 
 plants and produced by insects of different genera show great 
 differences. This seems to indicate that the stimulus of a partic- 
 ular genus of insect is given to a particular part of the host plant 
 or is of a peculiar kind, characteristic of that genus. However, 
 if the stimulus of two different genera of insects be applied to the 
 same part of the plant the results may be similar. (See Part II. ) 
 
 4. Within each family we find certain morphological resem- 
 blances ; e. g. , Aphididae. 
 
 * Weisuiann, Essays on Heredity, Vol. I, p. 95. 
 
Plate iS. 
 
 Cook on Gaixs. 
 
272 The Ohio Naturalist [Vol. II, No. 7 r 
 
 5. The families show parallel lines of development from a low 
 form of gall structure up to a high form . e.g., Aphididae and 
 Cynipidae. 
 
 6. I am inclined to believe that the modification of the plant 
 tissue is purely mechanical. The loss of differentiation between 
 palisade and mesophyll and the closing up of the intercellular 
 spaces would be a natural result of rapid cell division. The 
 elongation of cells in certain directions would be a natural result 
 of mechanical tension arising from rapid growth. In the family 
 Aphididae where the gall is primarily a folding of the leaf the 
 elongation of the cells is parallel with the surface of the gall. In 
 those galls where the formation is a thickening of the leaf the 
 long axis of the cells is perpendicular to surface of the formation. 
 
 7. The presence of at least two zones, of which the inner may 
 be considered nutritive, is very common. 
 
 8. The formation of the gall is probably an effort on the part 
 of the plant to protect itself from an injury which is not sufficient 
 to cause death. Both Adler and Fockeu consider that after the 
 first stages of formation the gall becomes an independent organ- 
 ism growing upon the host plant. 
 
 9. Trichomes are far more prominent in galls produced by 
 mouth parts than in those produced by oviposition. 
 
 10. It appears from these studies that the histological charac- 
 ters of the gall will prove very important in determining the 
 characters of the species. 
 
 Part II. Apical Bud Galls. 
 
 In my third conclusion in the preceding paper I have expressed 
 a belief that galls produced by the same genus of insects show a 
 decided resemblance even though produced on widely different 
 plants. Furthermore, this similarity seemed to be due to the par- 
 ticular part of the host plant to which the stimulus was applied. 
 
 The following study of the apical bud galls seem to indicate 
 that when corresponding parts of different plants are stimulated 
 by insects of different genera that the galls produced have char- 
 acters in common. 
 
 The gall produced by Cecidomyia solidaginis J^w . (Fig. 31) is 
 merely a large bunch of leaves at the end of the stem of Solidago. 
 The cone-shaped gall of Cecidomyia salicis-strobiloides O . S. (Fig. 32) 
 at the tip of the twigs of Salix is a bunch of leaves reduced in 
 size and so compactly arranged as to produce the peculiar cone 
 effect. A further examination of these two galls shows that the 
 tips of the stems are enlarged and that the larval chamber is in 
 the apex. 
 
 A superficial examination of the gall of Callirhytis claxnda Fitch 
 (Fig- 33 a - b. c - d.) shows no resemblance to the preceding galls 
 except in location at the tip of the stem. The gall is apparently 
 
Ohio Naturalist. 
 
 Plate ig. 
 
 :MU ^ : ^M : 
 
 Cook on Galls. 
 
274 The Ohio Naturalist. [Vol. II, No. 7, 
 
 a mere enlargement of the tip of the stem, and containing one or 
 more larval chambers. Examination of section under a compound 
 microscope, however, reveals a condition similar to that described 
 for C. solidaginis and C. s.-strobiloides. Each larval chamber is 
 in reality the apex of a bud. The young leaves of the bud are 
 closely applied to each other and their structure unaffected by the 
 insect. As the gall developes the leaves do not unfold but assume 
 a corky texture and in the fully mature gall their identity is 
 almost lost. 
 
 It is very evident that the larval chamber occupies a correspond- 
 ing position in each of these galls. The insect prevents the 
 elongation of the stem, thus causing the leaves of the apical 
 bud to be bunched and reduced in size. The fact that the leaves 
 of the Solidago reach the greatest development and those of 
 the Quercus the least development is probably due to the char- 
 acter of the plants. Of these three plants the growth of the 
 Solidago is the most rapid while that of the Quercus is the slow- 
 est. In Solidago the rapid growth may be sufficient to overcome 
 the injury and cause the bunch of leaves ; in the Salix where the 
 growth is not so rapid the leaves are smaller and more compact ; 
 in the Quercus where the growth is slowest the bud never opens 
 but becomes corky and the leaves gradually lose their identity. 
 
 This work was pursued during the year 190 1-2 in the Zoolog- 
 ical Laboratory of the Ohio State University under the direction 
 of Professor Herbert Osborn to whom I am indebted for many 
 valuable suggestions. 
 
 LITERATURE. 
 
 Only those references which were especially useful in preparing 
 this paper are cited. 
 
 1. Adler, Hermann, M. D., " Ueber den Generations — wech- 
 sel der Eichen Gallwespen" Zeitschrift fur wissenschaftliche 
 
 Zoologie. Bd. 35. Leipzig "Alternating Generations, a 
 
 Study of Oak Galls and Gall Flies," translated by C. R. Stratton. 
 Clarendon Press, Oxford. 
 
 2. Ashmead, W. H. "A Bibliographical and Synonymical 
 Catalogue of the North American Cynipidae with descriptions of 
 new species." Transactions American Ent. Soc. Vol. XII, pp. 
 291-304. 
 
 3. Ashmead, W. H. " Synopsis of the North American Sub- 
 families and Genera of Cynipidae." Trans. Amer. Ent. Soc. Vol. 
 XIII, pp. 59-64. 
 
 4. Ashmead, W. H. " On the Cynipidous Galls of Florida 
 with descriptions of new species and Synopsis of the described 
 species of North America." Trans. Amer. Ent. Soc. Vol. XIV, 
 pp. 125-158. 
 
 5. Bassett, H. F. "Description of several supposed new 
 species of Cynips, with remarks on the formation of certain Galls. ' ' 
 Proc. Ent. Soe. of Phil. Vol. II. No. 3, pp. 323-333. 
 
Ohio Naturalist. 
 
 Plate 20. 
 
 
 'WS^gP*^ 
 
 Cook on Galls. 
 
276 The Ohio Naturalist. [Vol. II, No. 7, 
 
 6. Bassett, H. F. "Descriptions of several new species of 
 Cynips, and a new species of Diastrophus. " Proc. Ent. Soc. 
 Phil. Vol. Ill, pp. 679-691. 
 
 7. Beutenmueller, William. "Catalogue of Gall-producing 
 Insects found within Fifty Miles of New York City, with Descrip- 
 tions of their Galls, and of some new species." Am. Mu. of Nat. 
 Hist. Vol. IV. No. 1. Art. XV. pp. 245-268. 
 
 8. Fockeu, H. " Contributions a l'Histoire des Galls, Etude 
 Anatomique de quelques especes. " Imprimerie & Librarie Ca- 
 mille Robbe. Lille. 18S9. 
 
 9. Fockeu, H. " Recherches anatomiques sur les Galls." 
 Etude de quelques Dipterocecidies et Acarocecides." Imprimerie 
 Typographique & Lithographiqua le Bigot Freres. 1896. 
 
 10. Garmau, H. " The Phytopti and other Iujurious Plant 
 Mites." Twelfth Rep. of State Ent. on Noxious and Beneficial 
 Insects of the State of Illinois, pp. 123-143. 
 
 11. Osten-Sacken, Baron R. " On the Cynipidaeof the N. A. 
 Oaks and their Galls." Proc. Ent. Soc. of Phil. Vol. I. No. 3. 
 pp. 47-72. 
 
 12. Osten-Sacken, Baron R. "Additions and Corrections to 
 title 11." Proc. Ent. Soc. of Phil. Vol. I. No. 8. pp. 241-259. 
 
 13. Osten-Sacken, Baron R. " Contributions to the Natural 
 History of the Cynipidae of the U. S. and of their Galls." Proc. 
 Ent. Soc. of Phil. Vol. II. No. 1. pp. 33-49. id. Vol. IV. 
 PP- 33i-38o. 
 
 14. Osten-Sacken,- Baron R. "Two New North American 
 Cecidomyia." Proc. Ent. Soc. of Phil. Vol. VI. pp. 219-220. 
 
 15. Osten-Sacken, Baron R. "Biological Notes on Diptera 
 (Galls on Solidago)." Trans. Am. Ent. Soc. Vol. II. pp. 
 
 299-303- 
 
 16. Packard, A. S. " Insects Injurious to Forest and Shade 
 Trees." Fifth Rep. U. S. Ent. Com. 1890. 
 
 17. Pergande, Theo. "The Life History of Two Species of 
 Plant-Lice Inhabiting both Witch Hazel and Birch. ' ' Tech. Series. 
 No. 9. U. S. Dept. Agr., Div. of Ent., 1901. 
 
 18. Perris, E. " Galloides des Cecidomyies." Ann. de la 
 Entom. de France, 4 e series, t. X. 1870. 
 
 19. Riley, C. V. "The Grape Phylloxera, Phylloxera vas- 
 tatrix, Planchon." 3rd, 4th, 5th, 6th, 7th, 8th Annual Reports 
 of the Noxious and Beneficial and other Insects of the State of 
 Missouri. 
 
 20. Thomas, Cyrus. 8th Report of the State Ent. on the 
 Noxious and Beneficial Insects of the State of Illinois. 1879. 
 
 21. Walsh, B. D. "On the Genera of Aphididae found in 
 the U. S." Proc. Ent. Soc. of Phil. Vol. I. No. 9. pp. 294. 
 
 22. Walsh, B. D. " On Dimorphism of the Hymenopterous 
 genus, Cynips, with an Appendix, containing hints for a new 
 classification of Cynipidae and a list of Cynipidae, including 
 
Ohio Naturalist. 
 
 Plate 21. 
 
 Cook on Galls. 
 
278 The Ohio Naturalist. [Vol. II, No. 7, 
 
 descriptions of several species inhabiting the Oak Galls of Illinois. ' ' 
 Proc. Ent. Soc. of Phil. Vol. II. No. 4. pp. 443-500. 
 
 22. Walsh, B. D. "On Insects, Coleopterous, Hymenop- 
 terous, and Dipterous, inhabiting the Galls of certain species of 
 Willows." Proc. Ent. Soc. of Phil. Vol. III. pp. 543-644; 
 id. Vol. VI. pp. 223-288. 
 
 EXPLANATION OF PLATES. 
 
 In making the drawings, a Bausch and L,omb microscope and 
 camera lucida were used. Figs. 1-7 were made with i-inch ocular 
 and 1-5-inch objective. The diagrams of the galls were not made 
 upon a definite scale. All other drawings were made with i-inch 
 ocular and 2 3-inch objective. 
 
 eviations : e. — epidermis. 
 
 nu. 
 
 — nutritive zone. 
 
 end. — endodermis. 
 
 0. e. 
 
 — outer epidermis. 
 
 f. — fibro-vascular bundle. 
 
 P- 
 
 — protective zone. 
 
 1. c. — larval chamber. 
 
 pa. 
 
 —parenchyma. 
 
 Cross section of leaf of Hicoria ovata. 
 
 
 
 Salix cordata. 
 
 
 
 I. 
 
 2. 
 
 3. " " Vitis vulpina. 
 
 4. " " Ulmus americana. 
 
 5. " " Acer saccharinum. 
 
 6. " " Tilia americana. 
 
 7. " " Ouercus alba. 
 
 8. a. b. Phytoptus ulmi on Ulmus americana. 
 
 9. a. b. " abnormis on Tilia americana. 
 
 10. a. b. quadripes on Acer saccharinum. 
 
 11. a. b. " acericola " " 
 
 12. Schizoneura americana on Ulmus americana. 
 
 13. a. b. Colopha ulmicola on Ulmus americana. 
 
 14. a. b. Pemphigus ulmi-fusus on Ulmus americana. 
 
 15. a. b. Hormaphis Hamamelis on Hamamelis virginiana. 
 
 16. a. b. Phylloxera carya-avena on Hicoria ovata. 
 
 17. a. b. c. " " fallax " " 
 
 18. a. b. c. " " globuli " " 
 
 19. a. b. " " spinosa " " 
 
 20. a. b. " depressa " " 
 
 21. a. b. " vastatrix on Vitis vulpina. 
 
 22. a. b. c. d. Cecidomyia gleditsiae on Gleditschia triacanthos. 
 
 23. a. b. pilulae on Quercus alba. 
 
 24. a. b. verrucola on Tilia americana. 
 
 25. a. b. Neuroterus irregularis on Ouercus macrocarpa. 
 2.i. a. b. Callirhytis tumifica " alba. 
 
 27. a. b. c. Holcaspis centricola " palustris. 
 
 28. a. b. Amphibolips inanis " rubra. 
 
 29. a. b. c. Dryophanta palustris " palustris. 
 
 30. a. b. c. Callirhytis papillatus " sp. 
 
 31. Longitudinal section of Cecidomyia Solidaginis on Solidago. 
 
 32. " " salicis-strobiloides on Salix cordata. 
 
 33. Callirhytis clavula on Quercus alba. 
 
 a. Longitudinal section. 
 
 b. Cross section. 
 
 c. " " of leaf from b. 
 
 d. " " of larval chamber from b. 
 
 Note : — P. vastatrix was also collected on V. bicolor; C. pilulae was also 
 collected on Q. rubra and Q. palustris. 
 
GALLS AND INSECTS PRODUCING THEM. 
 
 Melville Thurston Cook. 
 
 Part III. Lateral Bud Galls. 
 
 In Part II of this series of papers I gave a discussion of apical 
 bud galls. The lateral bud galls differ from the apical only in 
 point of location ; therefore, this (Part III) may be considered a 
 continuation of Part II. There is, however, considerable differ- 
 ence in the galls dependent upon the order and genus to which 
 the insect belongs and to the part of the plant which is attacked 
 by the enemy. These differences may be summed up briefly as 
 follows : 
 
 ( i ) Affection of the tip of the stem causing it to remain in its 
 incipient condition and the leaves to remain aborted, instead of 
 lengthening. This is well illustrated by the apical bud galls of 
 Cccidomyia solidaginis Lw. on Solidago ; Cecidomyia salicis strobi- 
 loides O. S. on Salix ; and Callirhytis clamda Fitch on Quercus 
 alba. (Part II, Figs. 31, 32, 33.) In these cases we have two 
 orders of insects represented but producing similar galls : this, as 
 previously explained, is no doubt due to the fact that the insects 
 affect corresponding parts of the host plant. 
 
 (2) Affection of the tip of the bud causing it to remain short 
 but to become large and globular. This is well illustrated by 
 Hokaspis globulus Fitch (Fig. 34, a, b, c.) By collecting speci- 
 mens of this gall in April or early part of May it is easy to 
 demonstrate that the gall is in reality an enlargement of the stem 
 part of the bud. The insect evidently deposits the egg in the 
 apical part of the incipient stem. This causes the stem to enlarge, 
 forming a globular body, but to remain so short as to form a 
 sessile gall on the main stem. The bud scales are at first very 
 prominent but gradually shrivel up and are lost, leaving a naked, 
 
42o The Ohio Naturalist. [Vol. Ill, No. 7, 
 
 globular gall. At this late stage the only evidence that we have 
 of its bud origin is its location at the node of the main stem. 
 The transition from bud to gall occurs verj^ early, before there is 
 any differentiation of the parenchyma tissue ; examination of the 
 structure of the gall fails to show any stem characters but does 
 show the Cynipidous gall character described in Part I of this 
 series. 
 
 (3) The third type of the bud gall is illustrated in Andricus 
 seminator Harris (Figs. 35, a, b, and 36, a, b.) Ashmead* refers 
 to this as a flower gall. It is not difficult to demoustate that this 
 gall is a true, compound bud gall, but whether it is a flower or 
 leaf gall is not so easily determined. The strongest evidence of 
 its bud character is its location at the node of the stem and the 
 presence of the leaf scales at its base. The writer gathered and 
 dissected a large number of galls of various ages and is confident 
 that this is a true compound bud gall. In Figure 35 a, we have 
 a short twig with three buds, one of which was attacked by the 
 insect ; the other two buds remained unaffected. Around the 
 base of the gall are four well-defined bud scales. In Figure 35 
 1), two buds were affected ; one of these has been removed show- 
 ing the scar where it was attached and also exposing the back 
 side of the compound gall formed from the other bud. A great 
 many galls of various ages were dissected ; the younger ones 
 showing the bud scales and the older ones showing the well- 
 defined scars by which it was easy to trace the number of buds 
 affected. Careful observations were made in hopes of finding a 
 gall which would show whether this was a leaf or flower bud, but 
 without success. However, from a careful microscropic examin- 
 ation of a number of galls I am inclined to consider it a leaf bud, 
 in which each leaf becomes a single gall of the large cluster and 
 in which the incipient stem remains short. The microscopic 
 examination of the single galls ( Fig. 36, a, b) shows that each 
 gall contains at least one (and usually only one ) fibro- vascular 
 bundle which in most cases is very much atrophied and in some 
 cases so much reduced as to be very indistinct. The writer 
 considers the fibro- vascular bundle as the mid-rib of the modified 
 leaf and the cottony part of the gall as the mesophyll part of the 
 leaf. This gall does not show the four zones which are charac- 
 teristic of the cynipodous galls as pronounced as other galls 
 which we have examined, but this point will be discussed in a 
 later paper. 
 
 (4) The fourth type of gall is illustrated by a cecidomyid gall 
 ( Fig 37) found upon Acer negundo in which the bases of the 
 petioles of a number of leaves from the same bud are enlarged, 
 
 *Ashmead. Wm. H.: "On the Cynipidous Galls of Florida, with descriptions of new 
 species and synopses of the described species of North America." Trans. Anier. Knt. Sec. 
 Vol. XIV. pp: I2S-I2S 
 
May, 1903.] Galls and Insects Producing Them. 421 
 
 thus forming a bulb-like compound gall. On the inner surface 
 of the base of each petiole is a cavity containing the larva. The 
 stem remains short but the outer leaves are fully developed in 
 most cases. 
 
 (5) Pachypsylla celtidis-gemma Riley (Fig. 38) is evidently a bud 
 gall very similar to the preceding. Only advanced stages of this 
 gall were collected, and therefore its development could not be 
 observed. From the specimens collected it appeared that each 
 scale and undeveloped bud formed a pocket for the insect, there 
 being a single insect under each scale. 
 
 CONCLUSIONS. 
 
 Bud galls are subject to considerable variation due to the fact 
 that they are produced by insects of different orders and that 
 these insects attack different parts of the buds and different 
 tissues in these parts. In all cases except the fourth the demands 
 of the insect are so great as to cause a very pronounced change in 
 the bud. In the fourth the modifications are not so pronounced 
 as in the other four types. 
 
 Part IV. Stem Galls. 
 
 Stem galls, according to my definition, include only those galls 
 which cause a swelling of the stem and with the larva placed in 
 or near the center, thus affecting the stelar and fibro-vascular 
 parts of the stem. This definition ma}' not be as broad as it 
 should be, but I hesitate to make it include other forms until I 
 have had an opportunity to make a more careful examination of 
 the questionable forms. The fact that such galls as H. globulus 
 (Fig. 34, a, b, c), which is frequently mentioned as a stem gall,- 
 are in reality bud galls, leads me to be doubtful of the origin of 
 galls which have similar locations. Many of the so-called stem 
 galls may be in reality bud galls and this point can be determined 
 only by a study of their development and structure. 
 
 Some galls occur on both leaves and stem, but in these cases 
 the gall affects only the outer layers of the cells of very young 
 twigs and these cells at this time resemble the leaf cells in both 
 structure and functions. Phylloxera carya-spinosa Shimer (Part 
 I, Fig. 19) and Phylloxera caryae-caulis Fitch (referred to in 
 Part V) are good examples of leaf galls affecting stems. 
 
 The Tepidopterous galls are usually stem galls and may be 
 either solid or hollow and are most common on Solidago. In 
 studying such galls it is necessary to examine first a normal stem. 
 
 The stem of Solidago. (Fig. 39) shows the ordinary dicotyledo- 
 nous character. The epidermal cells (e p) are firm and rather 
 hard. Just below these cells is the parenchyma zone (pa) of 
 closely-fitted cells and few intercellular spaces. Below the par- 
 enchyma zone are the fibro-vascular bundles (p. v. b.), which 
 
422 The Ohio Naturalist. [Vol. Ill, No. 7, 
 
 contain a large amount of woody, fibrous tissue. Inside the zone 
 of fibro-vaseular bundles and forming the axis of the stem, is the 
 stelar (st) made up of large parenchyma cells. 
 
 In Tiypcta solidaginis (Fig. 40) a solid globular gall on the 
 stem of Solidago, we find the walls of the outer parenchymatous 
 cells much thickened and numerous large intercellular spaces 
 which are not characteristic of the unaffected stem ( Fig. 39). 
 The fibro- vascular bundles (f. v. b. ) are spread out and flattened, 
 the sclerenchyma tissue and tracheary tissue being reduced and 
 the fibrous tissue increased in amount. The parenchyma tissue 
 of the stelar (st) part of the gall is increased in amount and the 
 size of the cells reduced. This tissue is undoubtedly very active 
 and well supplied with nutrition for the larva. Throughout the 
 tissue are tubes (tu) lined with cells smaller than the parenchj-ma 
 cells, brown in color, and not affected by haematoxylin stain. 
 These tubes are usually associated with small bundles of fibrous 
 tissue and are probably important factors in the nutrition of the 
 larva. They were not found in sections of normal stem of corre- 
 sponding age. 
 
 In Gelechia gallae-solidaginu Fitch (Fig. 41) an elongated, 
 hollow gall on Solidago, we find the parenchymatous tissue (pa) 
 near the surface increased in amount, the cells larger and the 
 walls thicker than in an unaffected stem, but no intercellular 
 spaces such as are found in T. solidaginis. The fibro-vascular 
 bundles (f. v. b. ) undergo comparatively little change, becoming 
 slightly flattened and thinner and with a reduction of the firmer 
 fibrous tissue. The larva chamber (1. c.) of the gall is lined with 
 a few layers of small parenchymatous cells (st) and is the stelar 
 part of the stem. This parenchymatous tissue is udoubtedly 
 used for food. 
 
 In Cecidomyia rigidae O. S. (Fig. 42) an elongated, hollow gall 
 common on Salix discolor, usually near the tips of the twigs, we 
 find considerable modification of the normal stem structure. From 
 the examination of a number of specimens it is very clear that the 
 enlargement of the stem is due to two factors : the formation of 
 large intercellular spaces near the surface, similar to those in T. 
 solidaginis (Fig. 40), and the formation of the larval chamber 
 (1. c.) in the stelar part of the stem. The parenchymatous tissue 
 lining the chamber is made up of cells very much smaller than 
 those in an unaffected stem. 
 
 The Fepidopterous galls on the young stems of Acer neguudo 
 and Coleopterous galls on Rubus villosus were examined but no 
 new points presented. I was unable to secure satisfactory speci- 
 mens of stem galls of Cynipidae. 
 
 Although the study of stem galls was in many respects unsat- 
 isfactory, I feel justified in giving the following brief conclusions : 
 
May, 1903.] Galls and Insects Producing Them. 423 
 
 CONCLUSIONS. 
 
 1. Stem galls show less variations than any other group of 
 galls, although they may be produced by insects from widely 
 different orders. This is undoubtedly due to the fact that the 
 various insects attack corresponding parts of the host plants. In 
 proof of this fact, it will be noticed that all these insects deposit 
 the egg within the tissues of the host plant and not on the surface. 
 
 2. The galls in general show an increase of parenchyma below 
 the epidermis, either a thickening of cell walls or a development 
 of intercellular spaces, a flattening of the fibro-vascular bundles, 
 an increase of parenchyma tissue in stelar part of stem and a 
 decrease in size of same. 
 
 Part V. Development oe gales. 
 
 A very large amount of material was collected for this paper 
 and great difficulty was experienced in getting the extremely 
 young stages because of the fact that young specimens were diffi- 
 cult to recognize and identify. The material was carefully killed 
 in either Fleming's solution or chromo- acetic, passed through the 
 alcohols, imbedded in paraffin, sectioned on a Zimmerman micro- 
 tone and stained in haematoxyliu. 
 
 The galls will be considered in the same order as in Part I of 
 this series. A consideration of the leaf structure is unnecessary 
 since that was considered in Part I. 
 
 I. GALLS OF ACARINA. 
 
 Young galls of Phytoptus quad ri pes (Fig. 43), P. abnormis (Fig. 
 44), and P. acericola (Fig. 45) were studied, and all show the 
 same developmental characters. The leaf becomes slightly pitted 
 on one side (usually the lower) and a corresponding elevation is 
 formed on the upper surface. This gradually enlarges until the 
 more or less spherical gall is produced. In P. abnormis the 
 spherical gall soon assumed an elongated form. The character- 
 istic cell structure of the leaf is lost and the cells become very 
 irregular in shape. The elongated character of the cells just 
 beneath the outer epidermis appears at a later period of the devel- 
 opment. At first the inner surface of the gall is perfectly smooth, 
 but very soon masses of cells are formed and project into the 
 cavity (Figs. 43 and 45). At about the same time trichomes 
 begin to develop from the inner epidermis (Fig. 44) and project 
 into the cavity. These trichomes grow very rapidly and almost 
 fill the entire cavity. 
 
 In the very young galls no fibro-vascular bundles are formed, 
 but in the older galls small bundles of fibrous tissue are numerous. 
 
 The first effect of the insect attack is undoubtedly to cause an 
 increase in the number of cells, which is an effort on the part of 
 the plant to heal the wound produced by the repeated puncturing 
 
4^4 The Ohio Naturalist. [Vol. Ill, No. 7, 
 
 of the cells by the parasite. Since the parasite continues its 
 attack upon different cells and the plant makes the repeated effort 
 to heal the wound, we have the very active production of cells. 
 The parasite making its attack upon one side of the leaf, causes 
 the unequal growth resulting in a cavity. The increase in size 
 of the gall causes a different tension upon the inner and outer 
 surfaces and results in the elongation of cells near the outer sur- 
 face as described in Part I. 
 
 When the galls first appear they are single, but in a very short 
 time others are formed just outside the first, thus forming a 
 cluster. 
 
 In Erineum anomalum (Figs. 47, 48, a, b), occurring on leaves 
 and petioles of walnut, we find a condition similar to that of the 
 Phytoptus galls except that the parasite is on a free surface 
 instead of in a partly closed cavity. I was able to secure a very 
 complete series of this gall. The first indication of the gall on the 
 petiole or rib of a leaf is the increase in the amount of parenchyma 
 tissue between the epidermis and fibro-vascular bundles. The 
 physiological character of this tissue is also changed to some 
 degree, since the cells are not so easily stained with haematoxylin, 
 have rather thick walls, and contain a considerable quantity of 
 tannin. The epidermal cells now begin to form trichomes (Fig. 
 47). The parenchyma tissue and trichomes both increase in quan- 
 tity, the walls of the cells become thinner (Fig. 48, a, b), and the 
 deeper parenchyma tissue gradually loses its tannin, while the 
 outer cells retain it in great quantities. 
 
 These galls always occur over a fibro-vascular bundle and are 
 apparently closely associated with them. These bundles become 
 modified to some extent. 
 
 The origin and development of these galls is the same as in the 
 Phytoptus galls except that the parasite works upon the exposed 
 surface instead of in a cavity. The fact that one produces a 
 cavity lined with trichomes while the other produces a protuber- 
 ance covered with trichomes, is probably due to the fact that the 
 latter is so closely associated with the fibro-vascular bundle which 
 prevents the curvature but causes the rapidly-formed cells to 
 swell outward into a protuberance. 
 
 2. GALLS OF THE APHIDIDAK. 
 
 In the Aphididae galls we have a condition very similar to that 
 just described for the Acarina galls except that the shape of the 
 galls are far more definite and they show a higher degree of 
 development. Trichomes are not so numerous and masses of cells 
 projecting into the larval chamber as described for Phytoptus 
 galls are very rare. In the youngest galls the cell structure of 
 the leaf is modified, resulting in the formation of a large number 
 of small, irregular cells, the same as in the Acarina galls. As the 
 
May, 1903.] Galls and Insects Producing Them. 425 
 
 galls grow older the cells near the outer epidermis become elon- 
 gated as in the Phytoptus galls. 
 
 In Pemphigzk ulmifusus (Walsh) Oestland (Fig. 49, a, b) on 
 U. Americana, we have the gall originating first as a fold in the 
 leaf which becomes developed into a conical structure. The struc- 
 ture of the gall shows that the characteristic structure of the leaf 
 is at first modified into a large number of small, irregular-shaped 
 cells (Fig. 49, b). The tendency for the cells near the outer 
 surface to elongate parallel to the surface begins with the further 
 development of the gall. In the very young galls the tannin is 
 in very small quantities, but increases as the gall grows older. 
 
 In Colopka ulmicola Fitch (Fig. 50, a, b) we have a condition 
 almost identical with P. ulmi-fusus. The gall first appears as a 
 slight fold in the leaf and later develops into the characteristic 
 cockscomb gall. The cell structure is the same as in P. ulmi- 
 fusus. 
 
 In Phylloxera carya-fallax Riley (Figs. 51, 52) on H. ovata, I 
 secured the youngest galls possible to detect and identify. These 
 galls showed a slight projection from both surfaces of the leaf, 
 but at first the gall was not so conical as at a later period of its 
 development. However, the youngest galls showed the charac- 
 teristic structure described in Part I of this series. The first 
 effect of the parasite attack appears to be the formation of a large 
 number of irregular cells. The arrangement of these cells is the 
 same in the young gall as in the more mature, but the fibro-vas- 
 cular bundles of the older specimens were not observed in the 
 young galls. 
 
 I was not so successful in securing young specimens of P. 
 c.-globuli Walsh (Fig. 53), but, so far as I was able to observe, 
 the line of development coincided with P. c.-fallax. However, 
 the upper wall of the gall is at first very thin and grows in thick- 
 ness as the gall approaches maturity. 
 
 Phylloxera ea>ya-eaidis Fitch of Hickory ovata was studied very 
 carefully from a very complete series of specimens. The material, 
 especially the younger galls, did not cut well, and so was not 
 satisfactory for drawings. However, the development and struc- 
 ture were of the typical Phylloxera type corresponding very 
 closely with that just described for P. c.-fallax. The only marked 
 peculiarity was the close association with fibro-vascular bundles, 
 the galls always occurring on very young green twigs, on mid-rib 
 or on prominent veins of the leaf. 
 
 Pemphigus populi-transversus Riley (Figs. 55, a, b, and 56, a, b) 
 and P. p.-caulis Fitch (Figs. 57, a, b, c, and 58, a, b, c) of the 
 Populus are galls growing on the petiole ; the former at some 
 point between the blade and stem, the latter at the base of the 
 leaf. In both cases the attack is made from the outside, the same 
 as in other Aphididae galls and in the Acarina galls. A careful 
 
426 The Ohio Naturalist. [Vol. Ill, No. 7, 
 
 study of an excellent series of both galls shows a cell structure 
 and development very similar to other Aphididae galls ; i. e., a 
 large number of small, irregular cells. In P. p.-transversus (Fig. 
 55, a, b) the gall originates as a swelling on the petiole and 
 within this swelling is a large cavity opening to the outside 
 through a slit. In the P. p.-caulis the same condition is true 
 but the attack of the insect causes a one-sided growth, resulting 
 in the petiole being twisted at right angles to the blade (Figs. 
 57, a, b, c, and 58, a, b, c). 
 
 A careful examination of the cell structure of P. p.-transversus 
 (Fig. 56, a, b) and a comparison with the unaffected petiole (Fig. 
 54, a, b) indicated a very rapid growth, resulting in the very 
 large number of small, irregular cells. The character of the 
 young and of the mature gall was practically the same, and not 
 different, as in the more highly developed galls of other orders. 
 The fibro-vascular bundles were very slightly affected. 
 
 P. p.-caulis showed the same cell structure and development, 
 and, judging from these points alone, one would be unable to 
 separate these two galls. 
 
 3. GALLS OF PSYLLIDAE. 
 
 In Pachypsylla celtidis-mamma Riley (Figs. 59 and 60, a, b, c) of 
 the Celtis occidentalis the youngest galls did not show a cavity, 
 but showed a modification of the leaf by which there is formed a 
 large number of small, irregular cells which can be readily sepa- 
 rated into two zones ; the upper made up of small, and the lower 
 of somewhat larger cells (Fig. 59). I was unable to secure speci- 
 mens intermediate between this stage and a later stage, showing 
 the true form of the gall (Fig. 60, a, b, c) The youngest galls, 
 showing the true form, exhibited four well-defined zones: (1) 
 epidermis, (2) zone of large, irregular-shaped cells, (3) zone of 
 elongated cells, (4) zone of irregular-shaped cells next to the 
 larval cavity. Adjacent to zone (3), but derived from zones (2) 
 and (4), are cells which even in very young galls show schleren- 
 chyma characteristics. As the gall approaches maturity this 
 tissue increases until in the mature gall it may be found in great 
 abundance. This gall is undoubtedly the most highly developed 
 of any of the Hemiptera galls which I have studied. 
 
 4. GALLS OF CECIDOMYIA 
 
 Although I have a large number of Cecidomyia leaf galls, I 
 have succeeded in getting a series of only two species. Since the 
 Cecidomyia show by far the greatest variation in structural char- 
 acters and the smallest number of typical group characters, two 
 species are not sufficient to draw a very definite conclusion. 
 
 In Cecidomyia gleditsiac O. S. (Fig. 61, a, b) the two halves of 
 the leaflet never have an opportunity to unfold, but there is a 
 
May, 1903.] Galls and Insects Producing Them. 427 
 
 growth of cells allowing the leaflet to enlarge and form the larval 
 chamber between the two halves. The cells are at first normal, 
 but gradually lengthen in an axis at right angles to the mid-rib. 
 This can be readily observed by comparing the section of the very 
 young gall (Fig. 61, a, b) with the section of the mature gall 
 (Part I, Fig. 22). 
 
 In Cecidomyia verrucola O. S. (Figs. 62 and 63) the youngest 
 showed a condition in which the mesophyll part of the leaf was 
 reduced or entirely removed by the larva. The upper epidermis 
 and palisade cells, the lower epidermis and cells next to it, form 
 the upper and lower walls of the larval chamber while the inter- 
 mediate mesophyll is removed. The inner layers of cells, i. e., 
 the cells next to the larval chamber, now grow and divide very 
 rapidly, gradually filling almost the entire cavity and reducing 
 the size of the chamber (Part I, Fig. 24). At the same time the 
 gall is increasing rapidly in size. 
 
 5. GALLS OF THE CYNIPIDAE. 
 
 Although a large amount of material was collected, only three 
 species were sufficiently complete to enable a satisfactory study. 
 However, several mature galls of species not described in Part I 
 of this series were examined, and all agreed with the statements 
 made concerning the general structural character of this group 
 of galls. 
 
 CallirkyHs papillatus O. S. (Fig. 64) was especiallj- difficult to 
 collect because of its very small size and close resemblance in 
 external appearance to other small Cynipidous galls. Examina- 
 tion of young Cynipidous forms, which I am reasonably certain 
 belong to this species, show all the zones in contact (Fig. 64). 
 As the gall develops the protective zones and parenchyma zones 
 separate but remain connected by elongated parenchymatous cells 
 (Parti, Fig. 30). 
 
 Dryophanta palustria O. S. (Fig. 65, a, b) appears as the leaves 
 unfold from the bud. The youngest galls collected were not over 
 two millimeters in diameter but showed the four zones well devel- 
 oped, with the second and third zones in contact, thus verifying 
 the views expressed in Part I. The cells of the innermost, or 
 nutritive, zone were large and very granular. Evidently this zone 
 was almost completely reduced by the larva in the specimen from 
 which Fig. 29 of Part I was drawn. In the next, or protective, 
 zone the cell walls were very thick. In the parenchyma zone the 
 innermost cells were small and numerous and the walls were thin, 
 and in both cases the long axis of the cells were at right angles 
 to the surface of the gall. As the gall grows older the intercellu- 
 lar spaces may become prominent among the cells of the paren- 
 chyma zone (Fig. 65, b). Careful examination of a large number 
 of specimens gave conclusive proof that the separation occurs 
 
428 The Ohio Naturalist. [Vol. Ill, No. 7, 
 
 between the protective and parenchyma zones, thus leaving the 
 two inner zones as a small sphere rolling free within the larger 
 sphere which is formed by the two outer zones. 
 
 In Diastrpphus siminis Basset (Figs. 66, a, b; 67; 68, a, b, c, d; 
 69 ) we have a Cynipiclous gall occurring on Nepeta glechoma. I 
 secured a very complete series of this gall and made a very careful 
 study of its development. In the youngest gall (Fig. 66,^, b) 
 we have the cell character of the leaf transformed into a mass of 
 small, irregular cells which can be readily divided into two zones, 
 the outer of which has the larger cells. At this time the cells are 
 very compact, but as the gall grows older intercellular spaces are 
 developed, the entire structure becomes loose and spongy and the 
 cells become larger. 
 
 As the galls grow older a well-defined zone of flattened cells 
 is developed in the parenchyma near the epidermis, and fibro-vas- 
 cular bundles ( f . v. b. ) are developed at right angles to the sur- 
 face (Fig. 67). Up to this time the cells are small, irregular and 
 compact. The epidermis (ep) and parenchyma (pa) zones are 
 well defined, but the distinctiion between protective and nutritive 
 zones cannot be made. 
 
 As the gall grows older a cleavage plane is formed in the paren- 
 chyma just inside the zone of flattened cells (Fig. 68, a). A 
 careful examination of the parts thus cut off and surrounding the 
 larval chamber (1. c. ) shows two well-defined zones which corre- 
 spond to the nutritive and protective zones described in Part I. 
 At this time there is no marked difference in the amount of food 
 supply of the tw r o zones. In the outer part formed by this cleav- 
 age plane we have the parenchyma (pa) and epidermal (ep) 
 zones (Fig. 68, c). Connecting the parenchyma and protective 
 zones we find fibro-vascular bundles ( f . v. b. ) surrounded by par- 
 enchyma cells (Fig. 68, d). The character of these connecting 
 strands is very similar to that described for H. centricola (Part I, 
 Fig. 27) and A. inanis (Part I, Fig. 28), but contains more par- 
 enchyma tissue than either. However, the parenchyma cells are 
 not so elongated as in C. papillatus (Part I, Fig. 30). As the 
 gall grows older the cells of the protective zone become clear and 
 the cell walls of the nutritive zone gradually thicken (Fig. 69), 
 many undergoing complete degeneration, while others assume the 
 character of the sclerenchyma. 
 
 CONCLUSIONS. 
 
 1 . All conclusions given in Part I are emphasized by the study 
 of the development of the galls. 
 
 2. In the formation of all leaf galls except the Cecidomyia 
 galls, the normal cell structure of the leaf is first modified by the 
 formation of a large number of small, compact, irregular-shaped 
 cells. In the galls of Acarina and Aphididae this is followed by 
 
May, 1903.] Galls and Insects Producing Them. 429- 
 
 a development of trichomes, especially the former. In all galls 
 the mesophyll is subject to the greatest modification. Many 
 small fibro vascular bundles are formed in this modified mesophyll. 
 
 3. The Acarin may be considered the lowest group of galls, 
 the Aphidid the next higher, the Cecidomyia galls the next 
 higher, and the Cynipidous galls the highest. However, many 
 of the Cecidomyia galls are lower than the Aphidid galls. 
 
 4. The galls of Acarina and Aphididae show the greatest 
 resemblance. In these cases the method of attack is very similar 
 and is first directed against the epidermal or adjacent layer of cells. 
 
 5. In some of the Cecidomyia galls (e. g. C. verrucola) the 
 larva appears to make its entrance into the mesophyll before there 
 is any pronounced modification of the cell structure. However, 
 the Cecidomyia galls are too varied and the study too incomplete 
 to make a positive conclusion. 
 
 6. Both Adler and Fockeu consider that after the first stages 
 of formation, the gall becomes an independent organism growing 
 upon the host plant. This is probably true in the highly devel- 
 oped galls of Aphididae, Cecidomyia and Cynipidae, but the 
 writer is very doubtful if this is true of the less complex galls of 
 Acarina, Aphididae and Cecidomyia. 
 
 This work was pursued during the year 1902-03, in the Biolog- 
 ical Laboratory of DePauw University, but was under the super- 
 vision of Professor Herbert Osborn, of the Ohio State University, 
 to whom I am indebted for many valuable suggestions. I am 
 also indebted to two of my former students, Miss S. Emma Hick- 
 man and Miss Margaretta S. Nutt, for aid in preparing slides and 
 making drawings. Drawings made by these two ladies are marked 
 with their initials. I also wish to express my thanks to my many 
 friends who have called my attention to, or have collected material 
 for, these investigations. 
 
 LITERATURE. 
 
 New literature will not be cited at this time, but a more com- 
 plete list will be given in connection with later papers upon this 
 subject. 
 
 EXPLANATION OF PLATES. 
 
 In making the drawings a Bausch & I,omb microscope, with 
 No. 2 ocular and }4 objective, and a B. & L camera lucida were 
 used. The drawings are, therefore, larger than those used in 
 Parts I and II, and the reduction not so great. The diagrams 
 are not made upon a definite scale. Drawings 34, a, b, c ; 35, a, 
 b; 37, 38, 55, a, b ; 57, a, b, c, and 58, a, b, c, were made from 
 nature, and are very little smaller than the original. The num- 
 bering of the drawings is continuous with Parts I and II. 
 
43° The Ohio Xaturalist. [Vol. Ill, No. 7, 
 
 ABBREVIATIONS. 
 
 ep. — epidermal zone. 
 
 pa — parenchyma zone. 
 
 pr. — protective zone. 
 
 nu. — nutritive zone. 
 
 f. v. b.— fibro-vascular bundles. 
 
 34, a. Bud of Hicoria ovata. 
 
 34. b. c. Holcaspis globulus on H. ovata. 
 
 35, a. Andricus semiuator gall and two buds on O. alba. 
 
 35. b. Andricus seminator gall and bud scar on Q. alba. 
 
 36, a, b. Section of Andricus seminator gall on Q. alba. 
 37- Cecidomyia gall on A. negundo. 
 
 38. Pachsylla c -gemma on C. occidentalis. 
 
 39. Cross section of stem of Solidago. 
 
 40. Trypeta solidaginis on Solidago. 
 
 41. Gelechia gallae solidaginis on Solidago. 
 
 42. Cecidomyia rigidae on Salix. 
 
 43. Phytoptus quadripes on A. saccharinum. 
 
 44- abnormis on T. Americanura. (Two larval chambers.) 
 
 45- " acericola on A. saccharinum. 
 46. Petiole of Juglans nigra. (Cross section.) 
 
 Erineum anomalum on J. nigra. Young gall 
 
 48, a. b. Erineum anomalum on J. nigra. (Mature gall.) 
 
 49, a, b. Pemphigus ulmi-fusus on U. Americana. 
 
 50, a, b. Colopha Ulmicola on U. Americana. 
 
 51, Phylloxera carya-fallax on H. ovata. 
 
 52, '• 
 
 53- carya-globuli on H. ovata. 
 
 54, a, b. Cross section of petiole of Populus monilifera. 
 
 35, a. Pemphigus populi-transversus on petiole of P. monilifera. (Yotmg gall.) 
 
 55, b. Same in section. 
 
 56, a. P. p.-transversus. Part of gall near opening into larval chamber. 
 
 56, b. P. p.-transversus. Section back of chamber and showing one fibro-vascular 
 
 bundle of the petiole. 
 
 57, a. P. p.-caulis. Young gall ; ventral surface. 
 57. b. Young gall; dorsal suiface 
 
 57. c- " Young gall ; open. 
 
 58, a. Ventral surface. 
 
 58, b. Dorsal surface. 
 
 Open. 
 
 59, Paehypsylla celtidis-mamma on C. occidentalis. (Young gall. 
 £0, a. P. c. -mamma. Diagram. 
 
 60, b. Section of dorsal part. (2 and 3.) 
 
 60, c. Section of ventral part. (3 and 4.) 
 
 61, a. b. Cecidomyia gleditsiae on G. triacauthos. 
 
 62, verrucola on T. Americana. 1 Young gall.) 
 63. 
 
 64. Callirhytis papillatus on Q. palustris. 
 
 65. a. b. Dryophanta palustris on O. palustris. 
 
 66. a, b. Diastrophus siminis on X. glechoma. 
 
 67. •• - •• ' •• 
 
 68. a. " " Diagram. 
 
 68, b. " Nutritive and protective zones. 
 
 68, c. Epidermal and parenchyma zones. 
 
 ■68, d. " Strand connecting protective and parenchyma zones. 
 
 •69- " Nutritive zone in gall almost mature. 
 
May, 1903.] Galls and Insects Producing Them. 
 
 Ohio Naturalist. 
 
 43i 
 Plate 13. 
 
 Cook on "Galls and Insects Producing Them. 
 
£* 
 
 Cook on "Galls and Insects Producing: Them. 
 
May, 1903.] Galls and Insects Producing Them 
 
 Ohio Naturalist 
 
 Cook on "Galls and Insects Producing Them." 
 
434 
 
 Ohio Naturalist. 
 
 The Ohio Naturalist. 
 
 [Vol. Ill, No. 7, 
 Plate 16. 
 
 Cook on " Galls and Insects Producing Them. 
 
May. 1903.] Galls and Insects Producing Them. 
 
 Ohio Naturalist. 
 
 Cook on "Galls and Insects Producing 1 Them. 
 
436 
 
 Ohio Naturalist. 
 
 The Ohio Naturalist. 
 
 [Vol. Ill, No. 7, 
 Plate iS. 
 
 Cook on -Galls and Insects Producing Them." 
 
[Reprinted from Ohio Naturalist, Vol. IV, No. 6.] 
 
 GALLS AND INSECTS PRODUCING THEM. 
 
 Melville Thurston Cook. 
 
 Part VI. Flower and Fruit Galls. 
 
 Galls affecting flowers and fruits are not so abundant as those 
 affecting leaves, but in many cases the insect which produces 
 flower or fruit galls also produces leaf galls. No sharp line of 
 distinction can be drawn between flower and fruit galls, since the 
 gall may form and mature without indication of fruit or ma3 T 
 form in the flower and mature as the fruit develops. Thus far I 
 have collected five species of flower and fruit galls representing 
 three orders of insects. 
 
 I. GALLS OF THE ACARINA. 
 
 Phytoptus sp. — on Euphorbia corallata L. (Figures 70 ; 71a, 
 b ; 72a, b). This mite produces galls on both leaf and flower. 
 The structure of the gall is the same in both cases and is identi- 
 cal with Phytoptus galls, previously described in Part I, (Figures 
 8-1 1). All my specimens of this gall were well advanced. The 
 structure of the leaf of E. corallata (Fig. 70) is typical. When 
 attacked by the Phytoptus the leaf becomes very much modified 
 by thickenings, ridges and convolutions (Figures 71a, b). The 
 palisade cells divide so that it is impossible to distinguish them 
 from the mesophyll, and the intercellular spaces are obliterated as 
 the result of the rapid cell division. The new cells are small and 
 very rich in protoplasm, but gradually become filled with tannin 
 as the gall approaches maturity. The tannin first forms in the 
 outer and most exposed cells of the gall while the inner layers of 
 cells retain their protoplasm very late. The Phytoptus restricts 
 its attacks to these inner and more protected parts. From a study 
 of these galls it is apparent that the Phytoptus is not working on 
 
 -Contributions from the Department of Zoology and Entomology, Ohio State Univer- 
 sity, under the direction of Prof. Herbert Osborn, No. 17. 
 
n6 The Ohio Naturalist. [Vol. IV, No. 6, 
 
 all parts of the gall at the same time, but gradually moves out- 
 ward over the surface of the leaf, thus increasing the size of the 
 gall and drawing its food supply from the newer part thus formed. 
 When the attack is made upon the flower we have a mass of 
 distorted tissue which is structurally the same as that produced 
 in the leaf gall (Figures 72a, b). The floral envelopes are the 
 first to suffer from the attack, the ovary with its contents is the 
 next greatest sufferer, while the stamens are frequently unaffected. 
 It is evident that the attack upon the flower must be made very 
 early in order to cause complete destruction. Very frequently 
 the floral envelopes will be very much deformed and the ovary 
 and the stamens very slightly affected. In other cases the ovary 
 will be very much enlarged and its chambers practically obliter- 
 ated. It is evident that the attack upon the ovary must be made 
 very earl}' to produce a great deformity. The partial immunity 
 of the stamens is probably due to their being very nearly mature 
 before the opening of the bud. 
 
 2. GALLS OF CECIDOMYIA. 
 
 Cecidomyia anthophila O. S. — on Solidago cauadense L. (Figs. 
 73a, b), makes the attack early and completely prevents the open- 
 ing of the bud. The gall is in the form of a hollow cone. The 
 transformation is so complete that the location is the only evidence 
 that the gall is produced from a flower bud. A section of the 
 gall shows the nutrient layers of the cells next to the larval 
 chambers, large parenchyma cells near the outer epidermis, and a 
 number of rather weak fibro- vascular bundles. 
 
 Cecidomyia sp. — on Ratibidapinnata Barnhart (Figs. 74a, b, c). 
 The entire bud is transformed into a gall with the larva in a 
 chamber in what was originally the ovary. All the floral parts 
 have become modified and united to form the gall. A section of 
 the gall (Fig. 74c) shows that the cells are more uniform in size 
 than in the preceding galls and that the fibro-vascular bundles 
 are practically obliterated. 
 
 Cecidomyia sp. — on Prunus virginiana L,. (Figs. 75a, b). My 
 specimens of this gall were mature. I am unable to say at what 
 time the gall originates, but it reaches its maturity with the fruit. 
 The gall is somewhat larger than the fruit, but otherwise resem- 
 bles it closely. The larva makes its exit through an opening at 
 one side of the stem. The larval chamber is very large, thus 
 giving the gall a bladder-like character. The cuticle is well 
 developed and the parenchyma cells below it are very large, while 
 the cells next to the larval chamber are much smaller. Weak 
 fibro-vascular bundles are also present. The wall of the gall 
 (Fig. 75b) is much thicker than the wall of the fruit at this time 
 (Fig. 75a), and parenchyma cells are much larger. The charac- 
 teristic stone (sclerenchyma) of the fruit is never developed in 
 the gall. 
 
April, 1904.] Galls and Insects Producing Them. 117 
 
 3. GALLS OF LEPIDOPTERA. 
 
 I gathered a number of Lepidopterous galls on Rudbeekia 
 laciniata L. which I was unable to determine. These galls occur 
 on both leaf and flower and are very large and fleshy. In fact 
 they were so fleshy and juicy that it was very difficult to secure 
 sections. The parenchyma cells were very large, and small fibro- 
 vascular bundles were numerous. The larval chambers were 
 numerous and each contained a single larva or pupa. In my 
 specimens the larvae were far advanced, many of them in the pupa 
 stage, but the cells next to the chambers were very rich in food 
 supply. 
 
 Part VII. Root Galls. 
 
 Amphibolips radicola Ashm. (Figs. 76a, b). — on Quercus alba 
 Iv. was the only root gall that I collected. The galls were borne 
 just under the surface of the ground at about the point of transi- 
 tion from stem to root. The)' were produced in great numbers 
 and so closely packed together as to assume the shape of figs. 
 Those nearest the surface of the ground and therefore slightly 
 exposed to the light were of a rich, red color, while those deeper 
 in the ground were almost white, slightly tinged with yellow. 
 Each gall contained from one to five larval chambers. The 
 younger galls showed four zones well defined (Fig. 76a). The 
 inner or nutritive zone was thick and the cells contained abund- 
 ance of protoplasm. The protective zone was thin and the cells 
 fibrous in character rather than sclerenchymatous. The paren- 
 chyma zone was thick and composed of large parenchyma cells. 
 The epidermal zone was relatively thick and the cells firm. As 
 the insects approach maturity the nutritive and protective zones 
 are entirely destroyed (Fig. 76b). The insect eventually makes 
 its escape through an opening in the side of the gall. 
 
 Part VIII. Histology of Galls. 
 
 Many of the histological characters of galls have been referred 
 to in the preceding parts. This part has been introduced at this 
 time for the purpose of adding a few additional facts which were 
 not clear at the time of the writing of the preceding parts. 
 
 A. Internal Stiucturcs. 
 
 I. galls of acarina. 
 
 These galls have been sufficiently discussed and need very little 
 attention at this time. In general these galls may be thrown 
 into three groups : (1) Those galls in which there is very little 
 distortion, but a modification of the epidermis, as in the case of 
 the Phytoptus on the beech ; (2) Convolutions of the parts as in 
 the case of P. ulmi (Fig. 8), P. abnormis (Figs. 9, 44), P. quad- 
 
n8 The Ohio Naturalist. [Vol. IV, No. 6, 
 
 ripes (Figs. 10, 43), and P. acericola (Figs, n, 45). These con- 
 volutions result in the formation of a more or less well defined 
 cavity, and trichomes are developed in great abundance in the 
 younger stages ; (3) Thickening of the parts which become cov- 
 ered with an abundant growth of trichomes as in the case of E. 
 anomalum (Figs. 47, 48). 
 
 The Phytoptus galls show two fairly well-defined zones, the 
 outer made up of rather large cells and the inner of much smaller 
 cells, which are very rich in protoplasm and which supply nour- 
 ishment for the young animal (Fig. 77). As the galls approach 
 maturity the protoplasm disappears, first from the outermost cells 
 and lastly from the cells on the inner surface. As the protoplasm 
 disappears the tannin accumulates in great abundance (Fig. 78). 
 
 2. GALLS OF THE APHIDIDAE. 
 
 Many of the Aphididae galls produce trichomes which soon 
 disappear. At first all the cells contain protoplasm and divide 
 rapidly, but as the galls approach maturity the tannin increases 
 in abundance. 
 
 Schizoneura americana Riley (Fig. 12), Colopha ulmicola Fitch 
 (Fig. 13), and Hormaphis hamamelis Fitch (Fig. 15) have been 
 considered in Part I. 
 
 In Pemphigus populi-transversus (Figs. 55, 56) and P. p.-caulis 
 (Figs. 57, 58) the thickness of the walls of the galls is much 
 greater than any other members of this famity and the cells are 
 more uniform in character. These galls are especially well sup- 
 plied with fibro-vasular bundles and are very dense. 
 
 In P. vagabundus (Fig. 112) we have a gall in which many of 
 the cells are elongated similar to C. ulmicola and H. hamamelis. 
 Its close structural resemblance to C. ulmicola and H. hamamelis 
 and unlikeness to P. p.-transversus and P. p.-caulis is due to the 
 fact that P. vagabundus, C. ulmicola, and H. hamamelis are 
 formed on the blades of the leaves, while P. p.-transversus and 
 P. p.-caulis are formed on the petioles which are made up largely 
 of fibro-vascular tissue. My specimens of these galls were mature, 
 and I am therefore unable to say anything concerning their early 
 stages. 
 
 In the Phylloxera galls all the cells are at first rich in pro- 
 toplasm and the tannin does not form in abundance until very 
 late. The two zones are. fairly prominent. In P. c.-caulus Fitch 
 on H. ovata, a gall which forms on both blade and petiole of the 
 leaf and also on young stems large intercellular spaces are formed 
 near the surface. 
 
 3. GALLS OF f-SVLLIDAE. 
 
 Pachypsylla c. -mamma Riley has been described in Part V 
 (Figs. 59, 60). 
 
April, 1904.] Galls and Insects Producing Them. 119 
 
 4. GALLS OF CECIDOMYIA. 
 
 These galls have been described in Part I (Figs. 22, 23, 24), 
 in Part V (Figs. 61, 62, 63), in Part VI (Figs. 73, 74, 75), and 
 in the Appendix (Figs. 114-119). In these galls the two zones 
 are usually fairly well defined, but the galls of this genus are so 
 different in character that it is difficult to give a definite descrip- 
 tion. The time for the formation of the tannin is variable, but 
 it is usually produced late and in great abundance. 
 
 5. GALLS OF THE CYNIPIDAE. 
 
 All these galls are very similar. The majority show the four 
 zones and in most cases these zones are well defined. The outer 
 zone is the epidermal which will be described later (Figs. 84-91). 
 The second is the parenchyma zone ; the third is the protective 
 zone made up largely of sclerenchyma, and the fourth or inner- 
 most is the nutritive zone. In many cases the second and third 
 zones become partially or entirely separated. This separation, 
 however, is not between the second and third zones as previously 
 stated by me in Parts I and V, and by Fockeu, but rather a sep- 
 aration of the tissues of the second or parenchyma zone, the 
 greater part of this zone clinging to the epidermal zone and a few 
 cells remaining attached to the protective zone. 
 
 Diastrophus siminis Bassett (Figs. 66-69) has been described 
 in Part V. The four zones are distinct and each shows the char- 
 acter previously referred to. 
 
 Diastrophus nebulosus O. S., described in the Appendix (Figs. 
 129a, b), is a stem gall in which the zones are well defined, the 
 protective zone being especially well developed. Each zone shows 
 the characters previously referred to. 
 
 In Amphibolips confluentus Harris (Figs. 121a, b, c) the first 
 and second zones are well developed, but the distinction between 
 the third and fourth is not so pronounced. 
 
 In Amphibolips inanis O. S. (Fig. 28) the four zones are well 
 defined. In the young gall (Fig. 79) the cells of the nutritive 
 zones are very rich in protoplasm and there is very little or no 
 distinction between the nutritive and the protective zone, but as 
 the galls approach maturity the cells of the protective zone 
 become very thick and are soon converted into sclerenchyma 
 (Fig. 80). 
 
 In Callirhytis papillatus O. S. we have the four zones well 
 defined (Fig. 30). As the gall approaches maturity the cells of 
 the nutritive zone lose their protoplasmic contents and become 
 very much shriveled, the protective zone is made up usually of 
 only two or three layers of cells. Next to the protective zone are 
 two or three layers of cells which are in reality a part of the 
 parenchyma zone. The large intercellular spacts formed in this 
 
i2o The Ohio Naturalist. [Vol. IV, No. 6, 
 
 zone are bridged by long unicellular threads, but no fibro-vascular 
 bundles (Fig. Si) 
 
 Dryophanta palustris O. S. galls show the four zones well 
 defined (Figs. 29, 65). When mature the contents of the cells 
 of the nutritive zone has been entirely used by the insect. The 
 protective zone consists of only two or three layers of sclerenchyma 
 cells, to which are attached a few cells of the parenchyma zone 
 (Fig. 82). 
 
 Andricus petiolicola Bassett (Fig. 124) produce a very hard 
 petiole or mid-rib gall which shows the four zones well defined. 
 There is no separation between the second and third zones. The 
 nutritive zone is at first very prominent, but it is reduced as the 
 gall approaches maturity. The protective zone developes its 
 sclerenchyma character rather late (Fig. 83) and gradually merges 
 into the two adjacent zones. 
 
 B. Epidermal Structures. 
 
 The epidermal cells vary in the size and in the thickness of the 
 cell walls. The galls may be smooth, pubescent or covered with 
 spiny structures. The amount of pubescence depends somewhat 
 on the natural pubesence of the host plant. Galls on such smooth 
 plants as Populus deltoides Marsh show very few and very small 
 trichomes, while galls on plants that are naturally pubescent are 
 likely to be pubescent. These trichomes vary in shape and gen- 
 eral character and are very prominent when the gall is young. 
 As the gall approaches maturity the trichomes usually disappear. 
 When these trichomes drop off their place of former attachment 
 is marked by a small mass of small cells, usually containing 
 tannin and from which imperfect rows of cells seem to radiate 
 (Figs. 84-90). 
 
 I. GALLS OF CYNIPIDAE. 
 
 Dryophanta palustris O. S. is very pubescent when young 
 (Fig. 84a). In the mature gall the cells are much larger, the 
 trichomes have disappeared and their point of attachment is made 
 visible by the accumulation of tannin (Fig. 84b). 
 
 All my specimens of Amphibolips inanis O. S. were fully 
 developed, but the points where the trichomes had evidently been 
 attached were very prominent (Fig. 85). These points are the 
 large, black spots so prominent on these large bladdery galls. 
 
 In Diastrophus siminis Bassett the trichomes are very large 
 (Fig. 86") and drop off very readily. 
 
 In Diastrophus potentillae Bassett the trichomes are very 
 numerous and each is at the apex of a very small elevation (Fig. 
 87). Examination of the epidermis of Acraspis eriuacei Walsh 
 show that its spines were due to similar but much more prominent 
 elevations. 
 
April, 1904.] Galls and Insects Producing Them. 
 
 2. GALLS OF THE APHIDIDAE. 
 
 Galls belonging to this family are usually less pubescent than 
 those belonging to the Cynipidae. The triehomes are usually 
 much shorter and frequently less numerous. Each trichome is 
 usually made up of a single cell (Fig. 88). The place where 
 these triehomes were attached is marked by an accumulation of 
 tannin, the same as in the Cynipidous galls (Figs. 89, 90). 
 
 Examination of the galls of the Phylloxera spinosa Shinier 
 show that the spines were due to the same cause as in the 
 Cynipidous galls (Fig. 87). 
 
 Galls of Pemphigus p.-transversus Riley (Fig. 91) and P. p.- 
 caulis Fitch were perfectly smooth, but the cell walls were much 
 thicker than in any other galls studied. 
 
 conclusion. 
 
 1. The inner layer of cells (i. e., those next to the larva) are 
 always supplied with nutriment until the insect is mature. 
 
 2. The development of the other layers of cells is for the pro- 
 tection of the larvae. These protective devices reach their highest 
 development in the Cynipidous galls. 
 
 3. In the very young galls there is usually little or no distinc- 
 tion between the nutritive and protective zones. The time of the 
 differentiation of the protective zones varies in different species. 
 
 4. The fibro- vascular bundles are most prominent in galls on 
 the petiole and mid-rib. 
 
 5. Most galls are covered with triehomes which disappear as 
 the galls approach maturity. The number of triehomes is varia- 
 ble in proportion to the pubescence of the host plant. 
 
 6. Spines are due to elevations composed almost entirely of 
 epidermal cells. 
 
 Part IX. Ovipositors and Mouthparts. 
 
 One of the most prominent questions concerning the formation 
 of galls which presents itself to the students of entomology and 
 botany and even to the most casual observer, is the exciting factor 
 in gall production. Is the stimulus from the ovipositor or mouth- 
 parts ? Is it mechanical or chemical ? The author believing that 
 the logical method of solving this problem was to first make a 
 careful study of the morphology and development of galls has 
 published the preceding parts of this paper. The author does 
 not claim to have found a complete solution of the problem, but 
 is hopeful that some of the facts stated in this series of papers 
 ma}- lead to more thorough and satisfactory studies of the prob- 
 lem. The problem presents many difficulties ; the parasites and 
 inquiliues which are usually present are frequently difficult to 
 distinguish from the real gall- maker ; this is especially true when 
 the study is confined to the larvae. In the following studies the 
 author is reasonably certain that the determinations are correct. 
 
The Ohio Naturalist. [Vol. IV, No. 6, 
 
 OVIPOSITORS. 
 
 Gall-making insects deposit their eggs by two methods, either 
 on the surface of the plant or within the tissues. Those insects 
 which deposit their eggs on the surface usually have mouthparts 
 developed for sucking, while those which deposit their eggs 
 within the tissues usually have mouthparts developed for biting. 
 Those which deposit their eggs on the surface of the plant are 
 the Acarina, the Hemiptera, and the Diptera. Those which 
 deposit their eggs within the tissues are the Hymenoptera and the 
 Lepidoptera. In this paper we have made a careful study of the 
 
 ovipositors of Cecidomyia gleditsiae, of Nematus sp — , Dry- 
 
 ophanta palustris, Amphibolips radicola, Andricus cornigerous, 
 A. seminator, and Rhodites radicum. A number of others were 
 examined, but because of the uncertainty as to determination are 
 not figured. 
 
 The Cecidomyia ovipositor (Fig. 92) is not suited to punctur- 
 ing tissues. The gall is never formed until after the hatching of 
 the larva. In this case it is evident that the stimulus, whether 
 mechanical or chemical, is produced by the larva. 
 
 Insects belonging to the genus Nematus deposit their eggs 
 either on the surface of the plant or in slits made by the ovipositor 
 (Figs. 93a, b). It is said that the galls are formed from these 
 wounds before the larva escapes from the egg, and in these cases 
 it is claimed that the irritating cause is a drop of fluid secreted 
 by the parent insect. Westwood claims that the egg increasing 
 in size is a result of imbibing sap from the wound in the plant. 
 It is well known that the eggs of some insects increase in size as 
 a result of the growth of the embryo within the egg. I have so 
 far been unable to make any satisfactory observations upon the 
 Nematus galls, but it is probable that the eggs increase in size 
 from the growth of the embryos and not as a result of the absorp- 
 tion of plant sap. It is also possible that the gall may be the 
 result of the mechanical irritation of the ovipositor or the enlarge- 
 ment of the egg or both. The wound caused by the ovipositor 
 of the Nematus is very much more severe than the wounds caused 
 by the ovipositors of the Cynipidous insects. 
 
 Adler, after a careful observation on Nematus Vallisnierii, says: 
 "This fly, which is armed with a finely serrated terebra, cuts 
 into the tender leaves of the end of the shoot of the Salix amyg- 
 dalina, and inserts her egg into the open wound, frequently plac- 
 ing several in the same leaf. At the same time the glandular 
 secretion flows into the wounded leaf. A few hours after this 
 injury the leaf surface presents an altered appearance, and new 7 
 cell formation begins freely, leading to a thickening of the sur- 
 rounding leaf surface. After the lapse of about fourteen days 
 the green and red-shaped gall is fully grown. If it be now 
 
April, 1904.] Galls and Insects Producing Them. 123 
 
 opened the egg can still be seen lying within the cavity. The 
 embryonic development is as yet unfinished and three weeks 
 elapse before the larva emerges from the egg to find around it the 
 material prepared for its nutriment. In this case the wound 
 caused by the fly is the immediate exciting cause of cell activity, 
 and leads to gall formation." 
 
 M. W. Beyerinck, in a paper regarding the growth of the gall 
 of Nematus caprea on Salix amygdalina holds a similar view. I 
 have not seen this paper, but an abstract* of it says: "The 
 production of the gall is undoubtedly due to the matter secreted 
 by the poison gland, which is, consequently, homologous with 
 the poison of Hymenoptera aculeata ; when the insect does not 
 deposit an egg in the wound which it makes, the quantity of 
 albuminous matter poured into the vesicle is always less than 
 when an egg is deposited ; by careful observation it is possible to 
 assure oneself that the size of the gall is always proportional to 
 the size of the wound and the quantity of albuminoid matter 
 introduced. By an experiment in which a deposited egg was 
 punctured by a fine needle, it was shown that the gall is due to 
 the parent and not to the egg ; but, of course, in such a case the 
 gall remains small ; neither the egg nor the larva are necessary 
 for its production, though their presence exercises a certain influ- 
 ence on the regularity of their development." 
 
 The ovipositors of the Cynipidae vary in length and in the 
 amount of coiling within the abdomen. All present the same 
 general characters. So far I have been unable to detect any 
 relationship between the length and character of the ovipositors 
 and the location and complexity of the galls (Figs. 94 to 98). 
 Adler claims that the egg is always deposited in or near the 
 Cambium layer of the plant. I am inclined to accept this state- 
 ment, but have made no special effort to verify it. If Adler' s 
 observations are correct the length of the ovipositor would be 
 associated not with the depth of the Cambium from the surface 
 of that part of the mature plant affected, but with the location 
 of the Cambium at the time of oviposition and with the difficul- 
 ties which the insect would experience in forcing the ovipositor 
 to the desired point. 
 
 Oviposition usually occurs before the buds are open, and the 
 eggs may be placed in three positions (1) in the stem, as in the 
 case of Rhodites radicum O. S., R. globulus Beut., Andricus 
 cornigerous O. S. ; (2) in the apex of the incipient stem as in 
 Andricus clavula Bassett, and Holeaspis globulus Fitch ; or (3) 
 in the leaves of the bud as in Rhodites bicolor Harris, Amphi- 
 bolips confluentus Harris, A. inanis O. S., A. ilicifoliae Bassett, 
 Neuroterus irregularis O. S., A. seminator, Callirhytis tumifica 
 
 'Jour. Roy. Micr. Soc, 1887, p. 746. 
 
124 The Ohio Naturalist. [Vol. IV, No. 6, 
 
 O. S., Holcaspis centricola O. S., Dryophanta palustris O. S., 
 and Callirhytis papillatus O. S. In these cases it is evident that 
 the force necessary to penetrate the bud may be as great or even 
 greater than the force necessary to penetrate a stem. Adler's 
 observations demonstrate that great force is used to penetrate the 
 buds and reach the desired point for depositing the eggs. 
 
 Beyerinck has demonstrated that the fluid ejected by the ovi- 
 positor of the Cynipidae is very different from the fluid ejected 
 from other Hymenopterous insects ; that it is without taste or 
 smell and does not irritate when injected under the skin. Adler 
 has demonstrated that this fluid cannot be considered as the stim- 
 ulus to gall production. It is probable that it may serve to attach 
 the eggs, or as an antiseptic, or as a seal for the wound. 
 
 Since the gall does not form until after the hatching of the 
 larva it is evident that oviposition does not furnish the stimulus 
 unless it may be that there is cell division but no swelling of the 
 plant tissues previous to the hatching of the larva. The author 
 has made no observations upon this point. Adler, in discussing 
 this question, says, in regard to Trigonaspis : " This fly pricks 
 the leaf in May, but months pass before any trace of gall forma- 
 tion can be seen. It has tolerably strong ovipositor with which 
 it cuts into the veins of the leaf, and in this way a distinct mark 
 is left wherever an egg has been inserted. Guided by these 
 marks it is easy to find the egg, but it is not until September that 
 the larva leaves the egg, and then gall formation begins." 
 
 MOUTH PARTS. 
 
 Since oviposition does not give an explanation of the stimulus 
 causing the formation of the gall it is necessar}' for us to turn our 
 attention to the mouthparts. 
 
 For convenience the insects may now be divided into two 
 groups, those with mouthparts for sucking, which make their 
 attacks upon the outside, and those with mouthparts for biting, 
 which make their attacks from the inside. Under the former are 
 included the Acarina, the Hemiptera and the Diptera ; under the 
 latter are included the L,epidoptera and the Hymenoptera. 
 
 I. HEMIPTERA. 
 
 The Hemipterous insects which produce galls may be placed in 
 the following order, with reference to the complexity of their 
 galls, beginning with the lowest : Schizoneura, Colopha, Horma- 
 phis, Phylloxera, Pemphigus and Pachypsylla. Mouthparts of 
 the following were carefully examined : Schizoneura americana 
 Riley, Colopha ulmicola Fitch (Fig. 99), Hormaphis hamamelis 
 Fitch, Phylloxera carya-fallax Riley, P. c.-globuli Walsh, P. 
 c.-spinosa Shinier, P. vastatrix Planchon, Pemphigus populi- 
 transversus Riley, P. p.-caulis Fitch, P. vagabundus Walsh, 
 
April, 1904.] Galls and Insects Producing Them. 125 
 
 Pachypsyllaceltidis mamma Rile)' (Figs. 100a, b), and P. c. -gemma 
 Riley. 
 
 The study of these mouthparts gave no new anatomical facts. 
 The different genera showed considerable variation as to length 
 of beak and setae. In general it may be said that the setae tend 
 to increase in the distance they may be protruded beyond the tip 
 of the beak as the galls approach complexity. This, however, 
 cannot be considered an exact rule, since the S. americana, C. 
 ulmicola and H. hamamelis have setae of practically the same 
 length, although the gall produced by S. americana is much 
 simpler than the galls produced by either C. ulmicola and H. 
 hamamelis (Part I, Figs. 12, 13 and 15). It was impossible to 
 make exact measurements of the distance the setae protruded 
 be3^ond the tip of the beak, since it was impossible to tell whether 
 the setae were full}' extended or partial^ retracted. The above 
 conclusions were reached after the examination of a large number 
 of specimens. 
 
 So far as I have been able to determine the insects do not 
 remain attached to any one point for a great length of time. The 
 P. c. -mamma (Figs. 100a, b) has a gall of the greatest complex- 
 ity, and the insect has setae which protrude farther beyond the 
 point of the beak than any other examined ; a large number of 
 these galls were opened and the position of the insect noted. The 
 insect was never found attached and apparently had no definite 
 point of attack. 
 
 The preceding observations emphasize Conclusions 6 and 8 of 
 Part I and a statement in the first of Part V. That is, the modi- 
 fication of the plant tissue to form the gall is purely mechanical, 
 being a continuous effort on the part of the plant to heal the 
 wound produced by the repeated puncturing of the cells by the 
 insect. When a branch is cut from a tree a growth is produced 
 which tends to cover the wound. In this case a single wound 
 and a single stimulus which is purely mechanical but which pro- 
 duces rapid growth for the purpose of covering the wound. In 
 the case of Aphididae and the Psyllidae galls the wounds are 
 more slight but repeated rapidly, the stimulus is mechanical and 
 the growth rapid, tending to cover the injury. 
 
 It is possible that the setae of the various genera may stimulate 
 different tissues and thus cause galls of varying complexity, but 
 upon this question I am not ready to give a definite statement. 
 
 2. DIPTERA. 
 
 The Cecidomyid galls occur upon a greater variety of hosts than 
 any other group of galls, and as previously stated in Part V, show 
 by far the greatest variation in structural characters and the 
 smallest number of typical characters. 
 
126 The Ohio Naturalist. [Vol. IV, No. 6, 
 
 The mouthparts of a number of larvae were examined (Figs. 
 101, 102), and all were practically the same; salivary or other 
 gland structures could not be demonstrated. 
 
 I am inclined to believe that the Cecidomyid galls are due to 
 purely mechanical stimuli and that the great variations are due 
 to the different tissues upon which the larvae feed. 
 
 Mr. W. A. Cannon,* in discussing a Cecidomyid gall on the 
 Monterey pine, says that the "larvae take their food only by 
 absorption through the surface of the body," also that " there is 
 no indication that the hypertrophy is either caused or affected by 
 any substance deposited with the eggs." 
 
 3. HYMENOPTERA. 
 
 We now come to the galls of greatest complexity and also to 
 those with which we have the greatest difficulty. These galls are 
 so very generally infested with parasites and inquilines that it is 
 difficult to decide which larva is the true gall producer. 
 
 A careful study of these shows that the insects have a very 
 strong pair of mandibles (Figs. 103 to 108), each working upon 
 two pivotal points. Some of these mandibles appear to have an 
 opening at the tip (Figs. 104, 105), and some showed what 
 appeared to be sacs or glands at the base (Figs. 104, 106b). In 
 one case at least (Fig. 104) these glandular sacs appeared to be 
 connected with the opening. The question that naturally pre- 
 sents itself is, are these openings for the purpose of pouring out 
 a fluid or are they suctorial as in the case of Chrysopa and other 
 families? In only two species was it possible to demonstrate 
 these structures. Some light is thrown upon this by Part VIII, 
 in which it was shown that the cell walls of the inner or nutritive 
 zones were not destroyed, but that the contents of the cells were 
 removed, causing them to shrivel. 
 
 The teeth of the mandibles are never on the same plane and 
 the mandibles become more and more chitinous as the larvae 
 approach maturity. The strength of the mandibles appears to 
 depend upon the density of the tissue through which the insect 
 works its way to the outside In A. inanis (104) and A. con- 
 fluentus (Fig. 105) the strength of the mandibles is practically 
 the same and the character of the galls very similar. In D. sim- 
 inis (Figs. 106a, b) the mandibles are stronger and the tissues of 
 the gall correspondingly denser. C. petiolicola (Fig. 103) is by 
 far the strongest of those studied, and the tissues through which 
 the insect must work its way the densest of the leaf galls (Fig. 
 124). 
 
 A study was made of the larvae from galls of C. papillatus. 
 This is a small, rather dense leaf gall. Farvae of two species 
 
 "Cannon, W. A. "The Gall of the Monterey Pine." The American Naturalist, Vol. 
 XXXIV, No. 406 (Oct., 1900), p. Soi. 
 
April, 1904.] Galls and Insects Producing Them. 1 2 7 
 
 were found (Figs. 107, 108). A careful study of the mouth- 
 parts lead me to consider No. 107 as a true gallmaker and No. 
 108 as a parasite. The mouthparts of the one which I consider 
 a true gallmaker were as strong as those of C. petiolicola (Fig. 
 103). The mandibles of the parasite (108) were equally strong 
 and showed what appeared to be rudimentary gland structures. 
 
 Holcaspis globulus Fitch was the only bud (i. e., incipient 
 stem gall, Part III, Fig. 34) gall examined. In the young larvae 
 the mouthparts are weak, but as the larvae approach maturity 
 the mandibles become very strong (Fig. 109) and well fitted to 
 cut the opening for the escape of the insect. However, the 
 mouthparts were not so strong as in the case of C. petiolicola, but 
 the gall of H. globulus is not so dense as the gall of C. petiolicola. 
 
 The mouthparts of Nematus poraum Walsh (Fig. no) were 
 very similar to those of the Cynipidae. I am not inclined to con- 
 sider the apparently glandular-like structure observed in a few 
 species of any great importance. They may be suctorial or they 
 may be degenerate organs. I consider the stimulus as purely 
 mechanical. The character of the gall may depend upon the 
 location, which would result in difference in tension in different 
 parts of the plant on which the gall may be located and also upon 
 the laws of natural selection, which will be considered in the latter 
 part of this paper. 
 
 It would be interesting to know the exact time that cell divi- 
 sion begins in the formation of a gall, but it is very difficult to 
 make satisfactory observations upon this point. Adler has made 
 successful observations upon this stage in Neuroterus laviusculus 
 and Biorhiza aptera. He says: "The moment the larva has 
 broken through the egg covering and has for the first time 
 wounded the surrounding cells with its delicate mandibles, a 
 rapid growth begins. This goes on so quickly that while the 
 posterior part of the larva is still within the covering a wall of 
 like growth of cells has already arisen in front. This rapid cell 
 increase can be easily explained because the irritation set up by 
 the emerging larva is exerted upon highly formative cells which 
 collectively possess every condition of growth. The cells which 
 are primarily around the larva cannot be distinguished from the 
 parenchymatous cells from which they proceed." 
 
 4. LEPIDOPTKRA. 
 
 A careful study was made of the mouthparts of the Gelechia 
 solidaginis Fitch (Fig. in) and upon an undetermined species 
 found upon Rudbeckia laciniata (Part VI). The mandibles are 
 larger and much stronger than in any of the Hymenopterous 
 galhnakers which I examined. The gall is also much stronger 
 than any of the Hymenopterous galls whose larvae were studied. 
 No glandular structures were observed. 
 
i28 The Ohio Naturalist. [Vol. IV, No. 6, 
 
 CONCLUSION. 
 
 i. The fluid secreted by the ovipositor is not an irritant, and 
 therefore cannot be the stimulus for gall production. 
 
 2. Since the gall does not form, excepting the Nematus galls, 
 until the appearance of the larvae, it is improbable if oviposition 
 is a stimulus for gall production ; and in those insects in which 
 the egg is not deposited within the tissues of the plant it is 
 impossible. 
 
 3. Glandular structures were observed in only a few of the 
 Hymenopterous larvae and these were of doubtful character. 
 
 4. Since it has so far been impossible to demonstrate the 
 presence of a chemical stimulus except in Nematus, we must 
 consider that the stimulus is usually mechanical. As previously 
 stated (Part I, Conclusion 3) the morphological characters of the 
 gall depend upon the genus of the insect producing it rather than 
 upon the plant upon which it is produced. The early history of 
 all galls except the Cecidomyid is practically the same ( Part V, 
 Con. 2). The shape and external character of the gall probably 
 depends upon the following : ( 1 ) The plant upon which the 
 attack is made; (2) Upon the part upon which the attack is 
 made ; (3) Upon the tissues affected ; (4) Upon possible results 
 of natural selection. 
 
 SUMMARY OF PARTS. 
 
 Next in importance to the problem of a stimulus giving rise to 
 a gall is the explanation of specific external characters. This 
 question is not easily answered and at the present time any 
 explanation must be largely theoretical. 
 
 The gall-producing insects are found in six orders, as follows : 
 1. Arachnida (mites); 2. Hemiptera (Aphidae and Psyllidae); 
 3. Diptera (Cecidomyidae and Trypetidae); 4. Hymenoptera 
 (Cynipidae and Tenthrenidae); 5. Uepidoptera, and 6, Coleop- 
 tera. The gall-producing habit must have originated independ- 
 ently in each of these orders and in some orders (Diptera and 
 Hymenoptera) it must have originated independently in each of 
 the two families represented. 
 
 The formation of the gall is due to two primary factors ; a 
 ' stimulus, usually mechanical, given by the insect, and nourish- 
 1 ment furnished by the plant. 
 
 Conclusions reached as results of previous studies and bearing 
 on this subject are as follows : 
 
 1. "Galls maybe classified into two general groups, viz.: 
 those produced by mouthparts and those produced by oviposition. 
 Those produced by oviposition may be considered the more highly 
 developed." (Part I, Con. 1.) 
 
April, 1904.] Galls and Insects Producing Them. 129 
 
 2. " The gall does not form until the appearance of the larvae. 
 Therefore all galls are produced by mouthparts." (Part VIII, 
 Con. 1.) The Nematus galls are an exception. 
 
 3. "The morphological character of the gall depends upon 
 the genus of the insect producing it rather than upon the plant 
 on which it is produced." (Part I, Con. 3.) 
 
 4. ' ' Within each family we find certain morphological resem- 
 blances." (Part I, Con. 4.) 
 
 5. "The families show parallel lines of development from a 
 low form of gall structure up to a high form." (Part I, Con. 5. ) 
 
 6. " The presence of at least two zones, of which the inner 
 may be considered nutritive." (Part I, Con. 7. ) 
 
 7. "The formation of the gall is probably an effort on the 
 part of the plant to protect itself from an injury which is not 
 sufficient to cause death. Both Adler and Fockeu consider that 
 after the first stages of formation the gall becomes an independ- 
 ent organism growing upon the host plant. This is probably true 
 in the highly developed galls of Aphididae, Cecidomyia, and 
 Cynipidae, but the writer is doubtful if this is true in the less 
 complex galls of Acarina, Aphididae and Cecidomyia." (Part I, 
 Con. 8 and Part V, Con. 6.) 
 
 8. "In the formation of all leaf galls except the Cecidomyia 
 galls the normal cell structure of the leaf is first modified by the 
 formation of a large number of small, compact, irregularly shaped 
 cells. In the galls of Acarina and Aphididae this is followed by 
 a development of trichomes, especially in the former. In all 
 galls the mesophyll is subject to the greatest modification. Many 
 small fibro-vascular bundles are formed in this modified meso- 
 phyll." (Part V, Con. 2.) 
 
 9. "Trichomes are far more common in galls produced by 
 mouthparts than in those produced by oviposition." (Part V, 
 Con. 9, and see Summary 2.) 
 
 10. "Variation in galls is due to their being produced by 
 insects of different orders, to their working upon different parts 
 of the plant and upon different tissues of these parts." (Part 
 III, Con., and Part IV, Con. 1.) 
 
 I. ARACHNIDA. 
 
 The Arachnida galls are of four types : (1) A modification in 
 the epidermis of the leaf as in the Phytoptus galls on maple and 
 elm ; (2) A fold in the plant tissue causing a cavity filled with 
 trichomes. among which the parasites live, as in the case of many 
 Phytoptidi (Figs. 8, 9, 10, 11, 43, 44, 45, Parts I and V) ; 
 (3) A swelling with an exposed surface covered with trichomes, 
 among which the parasites live, as in the case of Erineum 
 
i3° The Ohio Naturalist. [Vol. IV, No. 6, 
 
 anomalum (Part V, Figs. 47, 48); (4) The witehbroom forma- 
 tion, as in the case of the Phytoptus sp , and Sphaerotheca 
 
 phytoptophila Kell. and Sw. on Celtis occidentalis. 
 
 The author has studied only the second and third types. The 
 difference between these two may be accounted for by the fact that 
 the Phytoptus attacks the blade while the Erineum attacks the 
 petiole, mid-rib or larger vein. The part affected undergoes a 
 curvature in each case in the direction of the least resistance. 
 
 2. HEMIPTERA. 
 
 The method of attack by the Hemiptera is practically the same 
 as in Arachnida, i. e., by sucking mouthparts. The galls present 
 a complete serial line of development, the lowest form being a 
 simple curling of the leaf as in the case of Schizoneura americana, 
 the next higher, a simple folding of the leaf, as in the case of 
 Colopha ulmicola, the next higher is a more complex structure, 
 such as the Phylloxera galls and H. hamamelis, the next higher, 
 the slightly more complex, as in the case of the Pemphigus galls 
 (Figs. 12 to 21, and 49 to 58). The galls of the Pachypsylla 
 (Figs. 59, 60) are the most highl) 7 developed of the entire series. 
 
 Although in this case we have a complete series, it is difficult 
 to understand how this development has been produced. It may 
 be that the different forms are due to the attack being made upon 
 different tissues in each case, or to the degree in which the tissues 
 are injured. Upon this point we have no direct proof. However, 
 there is very little doubt that the stimulus is entirely mechanical. 
 
 3. DIPTERA. 
 
 As previously stated, the Cecidomyid galls are far more varied 
 in location and in morphological structure than any other group 
 of galls and show less number of characters peculiar to them- 
 selves alone. There is not sufficient data to draw even theoretical 
 conclusions concerning the influencing causes in their devel- 
 opment. 
 
 4. HYMENOPTERA. 
 
 As previously stated, the Cynipidous galls are the most highly 
 developed and show a greater number of morphological structures 
 peculiar to themselves than any other group (Part I, Con. 2 ; 
 Part V, Con 3). 
 
 Since the gall does not begin to develop until after the hatching 
 of the larvae, oviposition cannot be an important factor except in 
 so far as it is necessary to have the egg placed in certain tissues. 
 
 Examination of the mouthparts show few, small and insignifi- 
 cant gland-like structures the character of which is doubtful. It 
 is therefore probable that the stimulus is purely mechanical except 
 in the Nematus. But how are we to account for the s:reat mini- 
 
April, 1904.] Galls and Bisects Producing Them. 1 3 1 
 
 ber of specific external characters ? Fet us first review the struc- 
 tural characters of the leaf galls, since these galls show the most 
 uniform line of development. Considering Neuroterous irregu- 
 laris the gall of greatest simplicity, we can formulate the following 
 diagram : 
 
 — C. papillatus. 
 
 — A. confluentus. 
 -. — H. centricola. 
 - — A. inanis. 
 
 N. irregularis — C. tumifica. 
 
 -D. palustris. 
 -A. petiolicola. 
 
 Iii N. irregularis the zones are not so well developed as in 
 C. tumifica. In C. tumifica the zones are perfect, but in contact. 
 In C. papillatus the protective and parenchyma zones are sepa- 
 rated, but connected by long parenchyma cells. In H. centricola 
 and A. inanis the protective and parenchyma zones are connected 
 by fibro-vascular bundles. In A. confluentus they are connected 
 both by fibro-vascular bundles and by parenchyma cells (Fig. 
 121). In D. palustris the parenchyma and protective zones are 
 not connected. In A. petiolicola the zones are in contact, but 
 the tissues are very dense, due to location in the petiole of mid- 
 rib of the leaf. 
 
 If galls become independent structures they are undoubtedly 
 subject to the same laws of natural selection as any other group 
 of organisms, or if they be considered as parts of the plant they 
 must also be subject to the same laws of natural selection as any 
 other part of the plant on which they live. How, then, have 
 these laws affected the gall ? It may be a protective coloration 
 against birds and rodents, and other insects, but this cannot be 
 very important since man}- species of galls are very conspicuous. 
 Furthermore, animals make but very little use of galls for food. 
 So far I have observed other animals using galls for food but 
 once and then birds were tearing open the large galls of Pemphi- 
 gus vagabundus and eating the insects. The tannin which devel- 
 ops in such abundance in all galls as they approach maturity is 
 probably a great protection against insectivorous animals. 
 
 The greatest insect enemy with which the gall insect has to 
 contend is the great number of parasites. The size, shape and 
 character of the epidermal covering of the gall may be a protec- 
 tion against this numerous enemy. The thickness of the gall and 
 the density of the tissues, especially the protective zone, is an 
 
132 The Ohio Naturalist. [Vol. IV, No. 6, 
 
 important protective device. The large intercellular chambers in 
 the parenchyma zone place the larvae at a great distance from the 
 surface of the gall without increasing the amount of work neces- 
 sary for the mature insects to accomplish before reaching the 
 outside ; this is undoubtedly a great protection against parasites, 
 since it increases the difficulties for the parasite in reaching the 
 larvae with the ovipositor, The development of these protective 
 devices is probably the result of natural selection. Since the 
 character of the gall depends upon the insect, many variations in 
 the gall may also depend on variations in the stimuli given by the 
 insect. If these variations in character of epidermis, in thickness 
 of parenchyma zone, in the formation of large intercellular spaces, 
 in thickness and density of protective zone, are advantageous to 
 the insect in protecting it from the numerous parasites, these 
 characters may be perpetuated in succeeding generations and the 
 gall may increase in complexity. Natural selection is a reasona- 
 ble explanation. 
 
 It should be remembered that the plant is making an effort to 
 resist a parasite from which it cannot escape. The gall-maker 
 derives its nourishment without destroying its host and at the 
 same time strives to protect itself as far as possible from the great 
 number of parasitic enemies. The food supply first becomes a 
 part of the gall and upon this supply which, in the case of the 
 Cynipidae, is stored in the nutritive zone, it feeds. 
 
 Any irritation, such as the cutting or puncturing of plant tis- 
 sues, may and usually does cause excessive growth. It is proba- 
 ble that the primitive galls were of a type similar to the simplest 
 of the Phytoptus galls, i. e., a peculiar growth of the epidermal 
 cells. The next step in the evolution of the gall may be repre- 
 sented by a type similar to Schizoneura americana, in which case 
 the stimulus is greater, resulting in a curling of the leaf. The 
 next step may be represented by a type similar to the more com- 
 plex Phytoptus galls, H. hamamelis, C. ulmicola, the Phylloxera, 
 the Pemphigus and the most complex of the Pachypsylla galls in 
 which we find a series of more or less complex folds in the leaf up 
 to the increase in amount and differentiation of the tissue as in 
 the case of P. p. -mamma. 
 
 In the Cynipidous galls we have the greatest complexity, but 
 also a factor somewhat different from that in the forms to which 
 we have referred, i. e., the placing of the egg below the surface 
 and in those tissues upon which the larva is expected to feed. 
 It is impossible to say whether this habit of placing the egg below 
 the surface was acquired before or after the gall-making habit, 
 but it must be a great advantage to the insect. These galls, as 
 previously demonstrated, show the more complex serial line of 
 development of any of the galls, but even the simplest of these is 
 more complex than the most complex gall produced by any other 
 
April, 1904.] Galls and Insects Producing Them. 133 
 
 order of insect. This very complex development is due to an 
 early acquirement of the gall-making habit or to more rapid evo- 
 lutionary development as a result of the deposition of the egg 
 below the surface. 
 
 The greater part of the work connected with Part IX of this 
 series was conducted at the Eake Laboratory of the Ohio State 
 University at Sandusky, Ohio, and I am very much indebted to 
 the Director, Professor Herbert Osborn, for valuable assistance. 
 I also wish to express my thanks to the many friends who have 
 collected material and otherwise aided in these studies. 
 
 This series of papers will be presented to the Faculty of the College of 
 Arts, Philosophy and Science, of the Ohio State University, as the thesis 
 requirement for the degree of Doctor of Philosophy, June, 1904. 
 
 LITERATURE. 
 Continuous with the bibliography published with Parts I and II. 
 
 24. Adler, Dr. H. " Eege-Apparat und Eierlegen der Gal 1- 
 wespen." Deutsche Entomologische Zeitschrift XXI. 1877, 
 Heft II. 
 
 25. Beyerinck, Dr. M. W. " Beobachtungen uber die ersten 
 entwicklungsphasen einiger Cynipideugallen." Veroffentlicht 
 durch die Konigliche Akademie der Wissenschaften zu Amster- 
 dam, 1882. 
 
 26. Beijerinck, M. W. " Bydrage tot de Morphologie der 
 Plantegallen," 1877 
 
 27. Beijerinck, M. W. " De Gal van Cecidomyia aan Poa 
 nemoralis," Overgedrukt uit het Maandblad voor Natuurwet- 
 enschatten, 1884, No. 5. 
 
 28. Beijerinck, M. W. " Ueber Gallbildung und Genera- 
 tionswechsel bei Cynips calicis und uber die Circulansgalle," 
 Verhandelingen der Koninklijke Akademie van Wetenschappen 
 te Amsterdam, 1894. 
 
 29. Beijerinck, M. W. " Sur la Cecidiogenese et la Genera- 
 tion Alternante chez le Cynips calicis. Observations sur la galle 
 deL/Andricus circulans," Extrait des Archives Neerlandaises, 
 T. XXX, p. 387-444- 
 
 30. Busgeu, M. "Der Honigtau Biologische Studien an 
 Pflanzen und Pflanzenlausen. ' ' Zeitschrift fur Naturwissenschaft. 
 Bd. XXV. 
 
 31. Busgen, M. " Zur Biologie der Galle von Hormomyia 
 Fagi Htg. " Forstlich-naturwissenschaftliche Zeitschrift, Jan- 
 uar, 1895. 
 
 32. Courchet, M. L. " Etude sur less Galles Causees par des 
 Aphidiens." Akademie des Sciences et Eettres de Montpellier. 
 Memoires de la section des sciences. 
 
 33. Derbes, M. " Observations sur les Aphidiens qui fontles 
 Galles des Pistachiers." 
 
134 The Ohio Naturalist [Vol. IV, No. 6, 
 
 34. Eckstein, Dr. Karl. " Pflanzengallen und Gallenthiere. " 
 
 35. Fockeu, H. " Etude sur Quelques Galles. " Paris, 1897. 
 
 36. Heim, Dr. F. " Observations sur les Galles produites 
 sur Salix babylonica par Nematus salicis." 1893. 
 
 37. Kessler, Dr. H. F. " Die Entwicklungs und Eebens- 
 geschichte der Gallwespe Cynips calicis Brgst. und der von 
 derselben an den weiblichen Bluthen von Quercus pedunculata 
 Ehrh. horvorgerufenen Gallen, Knoppern genannt." Cassel, 
 1897. 
 
 38. Eacaze-Duthiers, M. Recherches pour servir a E'His- 
 toire des Galles." Extrait des Annals des Sciences Naturelles, 
 Tome XXIX. 
 
 39. Molliard, Marin. ' ' Recherches sur les Cecidies Florales." 
 
 1895. 
 
 40. Nabias, Dr. B. " Ees Galles et leurs Habitants." Paris, 
 
 1886. 
 
 41. Tschirch, A. " Ueber durch Astegopteryx, eine neue 
 Aphidengattung, erzeugte Zoo-cecidien auf Styrax Benzoin 
 Dryan." Deutschen Botanischen Gesell-schaft. 1890, Band 
 VIII, Heft 2. 
 
 42. Rubsaamen, Ew. H. " Uber Zoocecidien von der Balkan- 
 Halbinsel." 
 
 43. Paszlavszky, Jozsef. "A Rozagubacks Fejlodeserol." 
 Termeszetrajzi Fuzetek, Vol. V, Parte II-IV. Budapest, 1882. 
 
 44. Paszlavszky, Jozsef. " Beitrage zur Biologie der Cynipi- 
 den." Wiener Entomologische Zeitung II. 1883, Heft 6. 
 
 45. Prilleieux, M. Ed. " Etude sur la Formation et le Devel- 
 opment de Quelques Galles." 
 
 46. Pierre, M. l'abbe. " Ea Mercuriale et ses Galles." 
 Extrait de la Revue scientifique du Bourbonnais et du Centre de 
 la France, Juin, 1897. 
 
 EXPLANATION OF PLATES IX-XII. 
 
 The drawings were made with a Bausch & Lomb microscope. For Figs. 
 70-76 and Figs. 84-91 and Fig. 93b, a Number 2 ocular and ye objective. 
 For Figs. 77-83, a Number 2 ocular and ^ immersion objective. With 
 Figs. 92-98 and Figs. 106a, no and in, a }± ocular and 2 ; objective. For 
 Fig. 93 a Number 2 ocular and %' objective. The reduction is not so great 
 as in the preceding parts and therefore the figures are proportionately 
 slightly larger. The diagrams were not made upon a definite scale. The 
 numbering of the drawings is continuous with the preceding parts. 
 Abbreviations : e. epidermis. 1111. — nutritive zone. 
 
 ep. — epidermal zone. f. v. b. — fibro-vascular bundles, 
 
 pa. — parenchyma zone. 1. c. — larval chambers, 
 
 p. — protective zone. sc. — sclerenchyma. 
 
 FLOWER AND FRUIT GAIvLS. 
 70. Section of leaf of Euphorbia corollata. 
 
 71a. Diagram of section of Phytoptus sp gall on leaf of E. corollata. 
 
 71b. Section of 71a. 
 
 72a. Section of lower part of ovary of E. corollata affected by Phytoptus sp . 
 
April, 1904.] Galls and Insects Producing Them. 135 
 
 72b. Section of upper part of flower of E. corollata affected by Phytoptus sp . 
 
 73a. Diagram of cross section of Cecidomyid bud gall on Solidago canadense. 
 
 73b. Section of same. 
 
 74a. Diagram of longitudinal section of Cecidomyid gall on Ratibida pinnata. 
 
 74b. Diagram of longitudinal section of Cecidomyid gall on Ratibida pinnata. 
 
 74c. Section of 74b. 
 
 75a. Section of unaffected fruit of Primus virginiana. 
 
 75b. Section of Cecidomyid gall developed in fruit of P. virginiana. 
 
 ROOT GALL. 
 76a. Section of young gall of Amphibolips radicola. 
 76b. Section of mature gall of A. radicola. 
 
 HISTOLOGY. 
 
 77. Section of young gall of Phytoptus quadripes. 
 
 78. Section of young gall of Phytoptus abnormis. 
 
 79. Section of nutritive zone of young gall of Amphibolips inanis. 
 
 80. Section of mature gall of A. inanis. 
 
 81. Section of mature gall of Callirhyt is papillatus. (Nutritive, protective and part of 
 
 parenchyma zones.) 
 
 82. Section of mature gall of Dryophanta palustris. (Nutritive, protective and part of 
 
 parenchyma zones.) 
 
 83. Section of mature gall of Andricus petiolicola. 
 
 SURFACE SECTIONS OF 
 
 84. Dryophanta palustris. (Very young gall.) 
 84b. Dryophanta palustris. (Mature gall.) 
 
 Amphibolips inanis. 
 
 86. Diastrophus siminis. 
 
 87. Diastrophus potentilla. 
 Pachypsylla c.-mamma. 
 Colopha ulmicola. 
 
 90. Phylloxera c.-globuli. 
 
 91. Pemphigus p.-transversus. 
 
 OVIPOSITORS OF 
 
 92. Cecidomyia gleditsiae. 
 93a. Nernatus salicis-ovum. 
 93b. Nernatus salicis-ovum. 
 
 94. Dryophanta palustris. 
 
 95. Amphibolips radicola. 
 
 96. Andricus cornigerus. 
 
 97. Andricus seminator. 
 98a. Rhodites radicum. 
 98b. Rhodites radicum. 
 
 MOUTHPARTS OF 
 
 99. Colopha ulmicola. 
 
 100a. Pachypsylla c.-mamma, with setae extended. 
 
 100b. Pachypsylla c -mamma, with setae retracted. 
 
 101. Cecidomyia gleditsiae. 
 
 102. Cecidomyia pellex. 
 
 103. Andricus petiolicola. 
 
 104. Amphibolips inanis. 
 
 105. Amphibolips confluentus. 
 106a. Diastrophus siminis. 
 106b. Diastrophus siminis. 
 
 107. Callirhytis papillatus. 
 
 108. Parasite from gall of C. papillatus. 
 
 109. Holcaspis globulus, 
 no. Nernatus pomum. 
 
 in. Gelechia gallae-solidaginis. 
 
J 3 6 The Ohio Naturalist. 
 
 Ohio Naturalist. 
 
 [Vol. IV, No. 6, 
 
 Plate IX , 
 
 Cook on "Galls and Insects Producing Them. 
 
April, 1904.] Galls and Insects Producing Them 
 
 Ohio Naturalist 
 
 76' " V79~ 
 
 Cook on "Galls and Insects Producing Them." 
 
i38 
 
 The Ohio Naturalist. [Vol. IV, No. 6, 
 
 Ohio Naturalist. 
 
 Plate XI. 
 
 Cook on " Galls and Insects Producing Them." 
 
April, 1904.] Galls and Insects Producing Them. 139 
 
 Ohio Naturalist. p iate v// 
 
 Cook on "Galls and Insects Producing Them. 
 
140 The Ohio Naturalist. [Vol. IV, No. 6, 
 
 APPENDIX I. 
 
 GALLS AND INSECTS PRODUCING THEM. 
 
 Melville Thurston Cook. 
 Part I. Morphology of Leaf Galls. 
 
 I. GALLS OF THE APHIDIDAE. 
 
 The gall of Pemphigus vagabundus Walsh (Fig. 112) is evi- 
 dently formed as a result of the distortion of a large number of 
 bud leaves. My specimens of these galls were mature, so I was 
 unable to follow its development. Small fibro-vascular bundles 
 were numerous and tannin was formed in great abundance. The 
 structure was so modified that the leaf characters were lost ; the 
 cells were uniform in character, but were slightly smaller near 
 both the exterior and interior surfaces. 
 
 The galls of Pemphigus rhois Fitch (Fig. 113) are large, blad- 
 dery and evidently the pocketing of a single leaflet of the host 
 plant, Rhus glabra or R. typhina. My specimens of these galls 
 were fully mature, and I was therefore unable to follow the line 
 of development. The leaf structure was modified into the char- 
 acteristic Aphididae gall structure. Fibro-vascular bundles were 
 numerous and near the inner surface of the gall. Opposite each 
 bundle was a large cavity filled with some substance which I was 
 unable to determine. 
 
 2. GALLS OF CECIDOMYIDAE. 
 
 The galls of Cecidomyia pellex O. S. (Figs. 114a, b) are formed 
 by a thickening of the petiole, giving it the appearance of a long 
 fleshy bean pod with a slit along the upper side. This gall shows 
 three well defined zones ; an inner nutritive zone of small cells, a 
 parenchyma zone of larger cells and the epidermal zone. The 
 fibro-vascular bundles are numerous and are located between the 
 nutritive and protective zones and arranged around the larval 
 cavity and opening, the largest one just below the larval chamber 
 and corresponding to the mid-rib of the leaflet. 
 
 Cecidomyia impatientis O. S. (Fig, 115) is a fleshy gall occur- 
 ring on the leaves of Impatiens fulva. Some of my specimens 
 had the appearance of deformed flower buds, but upon this point 
 I was unable to decide. This gall showed two well defined 
 zones ; a zone of small cells lining the larval chamber and making 
 up about one half the thickness of the gall, and an outer zone of 
 large cells. Small fibro-vascular bundles were formed between 
 the zones. 
 
 The galls of Cecidomyia holotricha O. S. on Hicoria ovata 
 (Figs. 116a, b, c) are small and very firm. My specimens were 
 
April, 1904.] Galls and Insects Producing Them. 141 
 
 mature, but the cells lining the larval chamber were well supplied 
 with protoplasm, and numerous short trichomes were developed 
 from the dorsal surface and extended into the chamber. Tannin 
 was very abundant. 
 
 The gall of Cecidomyia tubicola O. S. on Hicoria ovata (Figs. 
 1 17a, b, c) is very similar to C. holotricha, except that the amount 
 of tannin is not so great. The upper wall of the gall is much 
 thicker than either the side or lower wall. The point of attach- 
 ment is not so large, but the gall is protected by a growth pro- 
 ducing a cup-shaped cavity in which the gall is developed (Fig. 
 117a). The inner layers of cells are very rich in protoplasm. 
 The cells are elongated in the long axis of the gall and fibro- 
 vascular bundles are more numerous than in C. holotricha, but 
 are very small. The cup-shaped structure (117c) in which the 
 gall is formed is composed of elongated cells. The palisade cells 
 in that part of the leaf opposite the gall are unaffected. 
 
 Cecidomyia viticola O. S. (Fig. 118) has the same general 
 character as C, tubicola, but is much longer. 
 
 Sciara ocellaris O. S. is one of the simplest of the Cecidomyidae 
 galls. The larva does not penetrate the tissues of the leaf, but 
 confines its attack to the outside, causing an indentation on one 
 surface of the leaf and a corresponding elevation on the opposite 
 surface (Fig. 119a) and also causing a very slight thickening. 
 The structure (Fig, 1 19c) when compared with that of the normal 
 leaf (Fig. 119b) shows the palisade transformed into ordinary 
 mesophyll and the intercellular spaces entirely obliterated. It 
 therefore corresponds in structure to the simple leaf-curl galls 
 produced by some of the Aphididae (e. g., Schizoneura Ameri- 
 cana Riley, Part 1, Fig. 12). 
 
 3. GALLS OF THE CYNIPIDAE. 
 
 My specimens of Rhodites bicolor Harris (Fig. 120) were well 
 developed when collected. I was therefore unable to determine 
 the early structural characters. The structure in these galls evi- 
 dently does not show the four well defined zones so characteristic 
 of this family. The inner cells are well supplied with nourish- 
 ment for the large number of larvae. 
 
 The galls of Amphibolips conflueutus Harris are very large and 
 have a single larval chamber in th</. center. The nutritive and 
 protective zones (Fig. 121a) can be distinguished, but are not so 
 well defined as in the closely related species, A, inanis (Part I, 
 Figs. 28a, b). The parenchyma and epidermal zones ( Fig, 121b) 
 are well defined and the space in the parenchyma is filled with a 
 cottony-like substance which upon close examination is composed 
 of fibro-vascular bundles (as in A. inanis, Figs. 28a, b, and H. 
 centricola, Figs. 27a, b, c) and of long, unicellular threads (Fig. 
 121c), as in C. papillatus (Figs. 30a, b, c and 81). 
 
142 The Ohio Naturalist. [Vol. IV, No. 6, 
 
 My specimens of Amphibolips illicifoliae Bassett were too far 
 advanced to admit of sectioning, but a careful examination indi- 
 cated that the zones were well defined and that the space in the 
 parenchyma zone is bridged by means of fibro- vascular bundles 
 as in A. inanis and H. centricola. 
 
 The galls of Amphibolips primus Walsh (Fig. 122) are very 
 firm and all the zones are well defined except the protective zone, 
 which is entirely absent. The parenchyma zone is very thick 
 and probably compensates for the lack of a protective zone. 
 There are very few small fibro- vascular bundles. 
 
 Galls of Amphibolips sculpta Bassett (Fig. 123) were more 
 succulent than other specimens which I have examined. My 
 specimens were mature, but the four zones were well defined. 
 The nutritive zone was almost obliterated, due to the age of the 
 gall. The protective zone was thin and the cell walls not very 
 thick. The parenchyma zone was very thick and composed of 
 large, succulent cells and was probably very important in furnish- 
 ing nutriment to the larva. Near the outer surface were numer- 
 ous small fibro-vascu'ar bundles. The epidermal zone was very 
 prominent and composed of small cells. 
 
 Andricus petiolicola Bassett is one of the firmest of the leaf 
 galls. It is formed either on the petiole or mid-rib and is com- 
 posed of very small, firm cells ( Fig. 124). The four zones are 
 well defined, but the protective zone is very thin and the cell 
 walls but very little thicker than in the neighboring cells The 
 parenchyma zone is very thick, composed of very small cells with 
 no intercellular spaces, but with many layers of long fibrous cells. 
 
 The galls of Acraspis erinacei Walsh (Fig. 125) are very con- 
 spicuous. The galls are always developed on the mid-rib of the 
 leaf, but contain no fibro- vascular bundles. The nutritive zone 
 is thick and very rich in protoplasm. The protective zone is 
 also thick and gradually merges into the parenchyma zone, which 
 is also thick. The epidermal zone is very irregular and is covered 
 with numerous unicellular trichomes. 
 
 The galls of Biorhiza forticornis Walsh are fig-shaped and the 
 larval chamber instead of being suspended in the center of the 
 gall, as is many others, is placed at the apex (Fig. 126a) and the 
 space between the protective and parenchyma zones, or rather in 
 the parenchyma zone, extends less than half way round the larval 
 chamber. My specimens were mature and I was unable to make 
 a careful study of the nutritive and protective zones. However, 
 the nutritive zone appeared to be relatively thicker, while the 
 protective zone was thin and merged gradually into the paren- 
 chyma zone ( Fig. 126b). The parenchyma zone was thick and 
 composed of large c^lls (Fig. 126c). Considerably more of this 
 zone remained attached to the protective zone than is the case 
 with most galls where this separation occurs. The cavity formed 
 
April, 1904.] Galls and Insects Producing Them. 143 
 
 by the separation of the cells in this zone is bridged by numerous 
 unicellular threads as in C. papillatus (Figs. 30a, b, c). In the 
 outer part of the parenchyma zone, but near the cavity, are 
 formed the fibro-vascular bundles. The epidermal zone is well 
 defined and the trichomes on the surface are uni-cellular (Fig. 
 126c). 
 
 4. GALLS OF TENTHREDINIDAE. 
 
 The galls of Nematus pomum Walsh were the only leaf galls of 
 this family that I secured and they were mature. There was no 
 indication of a zonal structure, but the cells were very uniform in 
 size and structure throughout the entire gall ( Fig. 127). Many 
 of the cells contained tannin and intercellular spaces were large 
 and evenly distributed. 
 
 Part II. Lateral Bud Galls. 
 
 Mature specimens of Holcaspis globulus Fitch show the four 
 well defined zones (Fig. 128). The inner nutritive zone is thick, 
 composed of small cells and well supplied with nutriment for the 
 larva. The protective zone is thin and composed of very small 
 cells with thin walls. It gradually merges into the nutritive zone 
 on the one side and the parenchyma zone on the other side. The 
 parenchyma zone is very thick, the cell walls medium in size and 
 the fibro-vascular bundles small and numerous. Further obser- 
 vations upon this gall emphasize the statement previously made 
 that it is the enlargement of an incipient stem. 
 
 Further observations upon the gall of Andricus seminator 
 Harris confirm the statement previously made that it is a com- 
 pound gall produced by the insect depositing an egg in each 
 element of the bud. 
 
 Part III. Stem Galls. 
 
 The gall of Diastrophus nebulosus O. S. (Fig. 129a, b) is a 
 very large swelling on the canes of Rubus villosus and is about 
 two or three inches in length. It contains a large number of 
 larval chambers each containing a single larva (Fig. 129a). The 
 four zones are especially well defined. The nutritive and protec- 
 tive zones are composed of a few layers of cells while the paren- 
 chyma zone is very thick, composed of smaller cells and more 
 dense than the corresponding zone in most galls of this family. 
 
 Andricus cornigerus O. S. (Fig. 130) produces one of the 
 hardest of the stem galls. My specimens of this were gathered 
 in the winter and were fully mature. The horn-like protuber- 
 ance is a closed tube extending to near the center of the gall. 
 This tube is composed of sclerenchyma tissue and evidently cor- 
 responds to the protective zone. Near the base of the tube is a 
 thin partition forming the larval chamber. When mature the 
 
144 The Ohio Naturalist. [Vol. IV, No. 6, 
 
 insect destroys this partition, travels to the end of the tube which 
 projects beyond the body of the gall, and there makes an opening 
 through either the end or the side of the tube and thus makes its 
 escape. Examination of young specimens would probably show 
 the four zones as well defined as in Diastrophus nebulosus. 
 
 Part IV. Development oe Galls. 
 
 Examination of very young specimens of Andricus seminator 
 Harris shows three well defined zones (Figs. 131a, b), the pro- 
 tective zone being undeveloped. The fibro- vascular bundles were 
 very numerous and distributed just beneath the epidermal zone. 
 I have examined a large number of these galls of various ages 
 and have been unable to find any trace of a protective zone. 
 Tannin develops in the outer cells very early and probably helps 
 to form a protection for the larva. 
 
 PLATES xin-xv. 
 
 112. Section of gall of Pemphigus vagabundus. 
 
 113. Section of gall of Pemphigus rhois. 
 114a. Diagram of gall of Cecidomyia pellex. 
 114b. Section of gall of Cecidomyia pellex. 
 
 115. Section of gall of Cecidomyia impatientis. 
 
 116a. Diagram of the gall of Cecidomyia holotricha. 
 
 116b. Section of the gall of Cecidomyia holotricha. 
 
 116c. Section of the gall of Cecidomyia holotricha. 
 
 117a. Diagram of the gall of Cecidomyia tubicola. 
 
 117b. Section of the gall of Cecidomyia tubicola. 
 
 117c. Section of the gall of Cecidomyia tubicola. 
 
 118. Diagram of the gall of Cecidomyia viticola. 
 
 119a. Diagram of the gall of Sciara ocellaris. 
 
 119b. Section of normal leaf of Maple. 
 
 119c. Section of gall of Sciara ocellaris. 
 
 120. Section of gall of Rhodites bicolor. 
 
 121a. Section of gall of Amphibolips confluentus. (Epidermal and parenchyma zones.) 
 
 121b. Section of the gall of Amphibolips confluentus. Nutritive and protective zones. ) 
 
 121C. Section of gall of Amphibolips confluentus. (Elongated cells in the cavity of the 
 parenchyma zone.) 
 
 122. Section of gall of Amphibolips piunus. 
 
 123. Section of gall of Amphibolips sculpta. 
 
 124. Section of gall of Andricus petiolicola. 
 
 125. Section of gall of Acraspis erinacei. 
 126a. Diagram of gall of Biorhiza forticornis. 
 
 126b. Section of gall of Biorhiza forticornis. (Nutritive and protective zones.) 
 
 126c. Section of the gall of Biorhiza forticornis. (Section of protective and epidermal 
 zones. ) 
 
 127. Section of the gall of Nematus pomum. 
 
 128. Section of the gall of Holcaspis globulus. 
 129a. Diagram of gall of Diastrophus nebulosus. 
 129b. Section of gall of Diastrophus nebulosus 
 130. Diagram of gall of Andricus cornigerus 
 
 131a. Diagram of cross section of gall of Andricus seminator. 
 131b. Section of young gall of Andricus seminator. 
 
April, 1904.] Galls and Insects Producing Them. 
 
 M5 
 
 Ohio Naturalist. 
 
 Plate XIII. 
 
 Cook on "Galls and Insects Producing Them. 
 
The Ohio Naturalist 
 
 [Vol. IV, No. 6, 
 Plate XIV. 
 
 Cook on "Galls and Insects Producing Them." 
 
April, 1904.] Galls and Insects Producing Them. 
 
 [ 47 
 
 Ohio Naturalist. 
 
 
 Cook on "Galls and Insects Producing Them." 
 
ft 
 
 s* 
 

 
y » < * °. 
 
 
 ■J>*r 
 
 <t 
 
 
 
 
 
 .^ 
 
 
 
 
 A 
 
 
 
 
 o - 
 
 ■ 
 
 i 
 
 ^ 
 *£ 
 
 ^ < 
 
 
 
 ^ J^ 
 
 
 ^ 
 
 A 9* 
 
 •> 
 
 "V 
 
 
 
 
 
 
 
 
 
 ^^. 
 
 ^ 
 
 

 
 
 
 
 
 
 
 
 
 
 
 
 **> 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 DOBBS BROS. 
 
 LIBRARY BINDING 
 
 ST. AUGUSTINE 
 /^S^v FLA. 
 
 32084 31