AG ALBERT R. MANN LIBRARY New YorkK STATE COLLEGES OF AGRICULTURE AND HoME ECONomIcs AT CORNELL UNIVERSITY Cornell University Library QK 605.E61T36 Ge ' es Wain 3 1924 001 725 724 mann MEMOIRS OF THE BOSTON SOCIETY OF NATURAL HISTORY; VOLUME IV, NUMBER VI. a THE ENTOMOPHTHOREAK OF THE UNITED STATES. By ROLAND THAXTER. a . BOSTON: PUBLISHED BY THE SOCIETY. APRIL, 1888. +p wre The undersigned is desirous of extending the present paper so as to include all the Entomogenous plants of North America, and wishes to procure by exchange or otherwise as large an amount of material for comparison as possible. For this purpose correspondence is invited, especially information concerning fungus epidemics among noxious insects. Address, after July 1, 1888, ROLAND THAXTER, CAMBRIDGE, APRIL, 1888. New Haven, Connecticut. VI. Tar ENTomMorpuTnoorREAE OF THE UNITED STATES. By Roranp THAXTER. Tue material upon which the following account is based was accumulated, for the most part, during the seasons of 1886-% from several localities in New England and in North Carolina, which were examined with such thoroughness as the limited time at my disposal would allow. The New England material was chiefly collected at Kittery Point, Maine, the southernmost point in the state, and in the vicinity of Boston; while the remainder is the result of two weeks botanizing in the alpine and sub-alpine region of Mt. Washington, N. H. The more southern forms represent three principal localities in or near the western portion of North Carolina. Of these Cullowhee, 2400 ft. above the sea, is the southernmost, having a flora of a distinctly southern type; while the two others, Cranberry (3250 ft.) and Burbank (EH. Tennessee, 3500 ft.), have a climate and flora not unlike that of the southern New England states. The eastern section of the United States is thus fairly well represented in so far as the localities which have been studied are concerned; yet it is scarcely necessary to remark that the forms obtained dur- ing a few weeks’ sojourn in each locality, in the course of general botanizing, can repre- sent only in a fragmentary way the Entomophthoreae of this section of the country. The forms occurring in the more remote regions of North America are, moreover, as yet al- most wholly unknown; and, although my observations have served to increase the num- ber of American representatives from four previously recorded forms to the considerable number hereafter enumerated, it cannot be supposed that the record is other than very imperfect. The present paper is therefore complete only in so far as I have endeavored to combine my own observations with those of previous students of the group in this country and in Hurope. For this purpose the literature of the subject has been consulted as far as has been practicable, and a list of the papers that I have myself seen is appended to this memoir. It should be understood, however, that this list is not intended as a complete record of all that has been written upon the subject, and is merely given for convenience of refer- ence in the text. The Russian publications of Sorokin were kindly procured for me in St. Petersburg by Mr. Charles Eliot, and for some knowledge of their contents I am indebted to Mr. Ivan Panin. For the privilege of examining the remaining papers, not contained in the Uni- versity libraries in Cambridge, together with other invaluable assistance, I am indebted to Professor Farlow, in whose laboratory the microscopic work upon my paper has been for the most part done. To Miss Hapgood I owe certain extracts from the Polish of MEMOIRS BOSTON SOC. NAT. HIST., VOL. IV. 20 (133) 134 ‘ ROLAND THAXTER ON THE Nowakowski, and I am also indebted to Professor Farlow, Messrs. C. V. Riley, L. O. Howard, Henry Edwards and A. F. Chatfield for several interesting specimens. To Mr. C. W. Woodworth and Prof. 8S. W. Williston I also owe certain entomological de- terminations. The plants that are to be considered in the present paper belong to a class which, al- though made up of several groups differing widely from one another in their habit and affinities, is yet, by reason of a peculiarity common to all its members, possessed of a certain individuality of its own that renders it susceptible of a consideration apart from all other forms of plant life. This peculiarity, by reason of which the class is usually characterized as entomogenous or entomophytous, consists in an obligatory parasitism upon insects, which, although in some instances it exists without apparent injury to the insect host, is usually of such a nature as to cause its death; often resulting, especially among’ noxious insects, in widespread mortality. Although a few of the more common or conspicuous forms of entomophytous plants had attracted the notice of botanists even in the last century, it was not until within comparatively recent years that they began to be studied with any care, and the work of Robin? is the first contribution of importance on the subject. This work still remains, with two exceptions,’ the only attempt that has been made to bring together all the known forms of insect parasites; but since its publication very important contributions have been made to our knowledge of the subject, through the medium of numerous scattered papers. My attention was first turned in this direction in the course of entomological studies on the life-histories of certain insects; in the course of which I was often greatly annoyed by losing large numbers of larvae and pupae through the agency of fungi. Having by this means and from other sources accumulated a certain amount of material, it was my first intention to include in my paper all the entomogenous plants recorded from Amer- ica; yet, owing to the many difficulties presented by the ascomycetous forms, involving a careful study and comparison of more abundant material than I could command, as well as by reason of the considerable additions to our Entomophthoreae resulting from my observations, I have decided to confine myself for the present to the members of this family, trusting to a future opportunity of extending my paper in conformity with my original plan. In the meantime a brief summary of the more important groups may not be out of place in this connection, and will be a fitting introduction to the more detailed consideration of the Entomophthoreae which follows. Summary of entomogenous plants.— Although the spiders and myriapods are not ex- empt from the attack of peculiar vegetable parasites, the hexapod insects offer by far the greater number of instances of this nature. Among the seven orders of the latter class usually enumerated, the Neuroptera and Orthoptera are almost wholly free from such attack; and, until recently, the first named order was considered wholly exempt in this respect. Of the remaining orders the Lepidoptera and Diptera are apparently the great- est sufferers; while the Hemiptera, Coleoptera and Hymenoptera are about equally af- 1C. Robin, J. ¢. that are known to form the bases of fungoid parasites, 2Sorokin, J. c., C, and Gray, G. R: Notices of insects London, 1858. ENTOMOPHTHOREAE OF THE UNITED STATES. 135 fected. Of the different stages of insects the imagines, larvae and pupae may ail be parasitized, and in some instances a single parasite may attack all these stages in one or more species of the same or different orders; while in others it may confine itself to a single stage or species. Entomogenous plants may in a general way be referred to five principal groups: one including the bacterial forms which produce disease in insects; a second represented by certain entophytous algae; and three others all belonging to the fungi proper. The first mentioned group, represented by the Bacteria, is chiefly of interest from an economic, rather than from a botanical point of view, as the supposed cause of destruc- tive epidemics among useful as well as noxious insects. Instances of this kind are pre- sented by the disease known as flacherie so destructive to silk worms, and in affections of a similar nature in other insects, where the “active principle” has, in some cases, been traced to bacterial forms which have been considered sufficiently peculiar to receive distinctive names. The systematic study of the group is necessarily one of great diffi- culty, and any opinion as to the validity of specific distinctions in such cases can only be formed by specialists in this department; but from a practical standpoint the existence of such affections promises to afford an important means of defence against noxious in- sects. The second group includes a small number of peculiar filamentous algae, represented by Enterobryus and its allies, that live attached to the digestive tracts of certain myriapods and coleopterous larvae. They are apparently nearly related to Oscillaria or Beggiatoa among the Protophytes; but, owing to insufficient observations upon them, their exact af- finities are unknown. ‘Their habit is probably one of commensalism, rather than of true parasitism; the partially digested food of the host being absorbed directly from the di- gestive tract." The fungoid parasites of insects are, as before mentioned, represented by three chief groups: the Entomophthoreae, the Laboulbeniaceae and the entomogenous forms which constitute the bulk of the genus Cordyceps. Since the first of these is to receive spe- cial consideration hereafter, it need only be said that its members are closely allied to the Mucorini among the Zygomycetes, and are entomogenous with few exceptions. The Laboulbeniaceae constitute a small group of very peculiar and minute forms which have been placed by DeBary among the doubtful Ascomycetes. Their parasitism is an external one, which apparently results in little if any inconvenience to the host; each individual being fixed by a pedicellate attachment to the legs, thorax or other portion of the affected insect. Several genera on Diptera, Coleoptera, etc., are described by Pey- ritsch? to whom we are principally indebted for our knowledge of the group, although the first genus of the family (Laboulbenia) was described and figured by Robin.2 The single American representative thus far recorded has been described by Professor Peck as Appendicularia entomophila, n.g. et sp.* The pyrenomycetous genus Cordyceps affords by far the most conspicuous examples of entomogenous plants, many of which are of large size, or brightly colored, and have 1See Leidy, Smiths. Contr. to Knowledge, v, pp. 1-67 p. 227 (1878): 72, 1, p. 377 (1875), plates. (1853) and Robin J. ¢., p. 395. 37. c., p. 622, plates. 2 Sitz. d. Akad. wiss. Wien., 64, 1, p. 441 (1871): 68, 1, 4 Peck, 38th Report, p. 95, with plate. 136 ROLAND THAXTER ON THE therefore received more attention from earlier botanists. Their imperfect or “Jsaria” condition is familiar to all entomologists as a pest in breeding cages and puparia; al- though the ascigerous condition is usually of rare occurrence. They attack all orders of hexapod insects, larvae and pupae as well as imagines, and also certain spiders ; often producing what is vulgarly known as a “vegetable sprout” several inches in length. A considerable number of American species are recorded; yet, owing to the lack of suffi- ciently well-marked microscopic characters, as well as to the scarcity of good material for study, the group presents many difficulties, as is usual in cases when too great re- liance has been placed upon gross characters as a means of specific distinction. In addition to-the groups above mentioned there are several other isolated instances of entomogenous fungi, among which should be mentioned the so-called Botrytis Bas- stana which produces the disease known as Muscardine, so destructive to silk worms in Europe and apparently identical with a similar form occurring in this country. Facultative parasites of insects——In addition to the obligate parasites briefly enumer- ated above, insects are often subject to the attack of numerous small moulds and bacteria which are in no sense peculiar to them, although they may temporarily assume a habit which is practically that of a true parasite, entering the living host and causing its death. It seems also probable that one or two forms which are truly entomogenous are yet saprophytes, as in the case of a certain Cordyceps (C. armeniaca) presumably growing upon the remains of insects in the excrement of insectivorous birds, as well as the members of the genus Basidiobolus hereafter mentioned, which occur upon the excre- ment of frogs and lizards. DeBary has also pointed out that the species of Cordyceps are normally partial saprophytes, since they attain their full development after the death of the host; but whether wholly parasites or saprophytes, or parasites and saprophytes combined, their peculiarity in growing naturally only upon insects or insect remains con- stitutes them entomogenous, in the sense in which I use the term, to the exclusion of such forms as Penicillium, Aspergillus, Cladosporium and the like; which, although they may at times not only grow on insects, but become temporarily truly parasitic upon them, are yet found in nature on a great variety of other substances. With this brief reference to entomogenous plants in general we may now turn to the consideration in detail of the group which forms the subject of the present paper. ENTOMOPHTHOREAE. General characters—This family at present comprises several genera, the members of which are not all entomogenous, though closely related structurally. “They are dis- tinguished by the production of numerous hyphae of large diameter and fatty contents, which, in the insect forms, ultimately emerge from the host in white masses of charac- teristic appearance and produce at their extremities large conidial spores which are vio- lently discharged into the air and propagate the disease. The common house-fly fungus is perhaps the most familiar example of the kind, and no one can have failed to notice the affected flies in autumn or late summer adhering to looking-glasses or window-panes surrounded by a smoky halo of discharged conidia. In addition to these conidia the propagation of the fungus, after long periods of rest, may be provided for by the forma- ENTOMOPHTHOREAE OF THE UNITED STATES. 137 tion of thick-walled resting spores adapted to withstand successfully the most unfavorable conditions. These resting spores, which may be either sexual (zygospores), or asexual (azygospores), finally germinate and .produce conidia that are discharged in the usual fashion and serve to infect fresh hosts. Such in brief is the general mode of develop- ment in Entomophthoreae; yet it is subject to so many variations and modifications in the different genera and species that a detailed comparison of them is instructive as well as necessary for a sufficient understanding of the family. I shall therefore consider each stage among the Hmpusae in some detail, having first briefly mentioned the more impor- tant points of structure in the remaining genera. These genera are four in number: Completoria, Conidiobolus, Basidiobolus and Massospora, the members of which, as al- ready mentioned, are not all entomogenous. The genus Completoria, which, as has been pointed out by Nowakowski and others, should be placed among the Entomophthoreae, was discovered by Lohde (J. ¢.) in the pro- thalli of ferns and has been subsequently more thoroughly investigated by Leitgeb (J. ¢.). Its presence is indicated by brown spots upon the prothallus within the cells of which it exists in the form of short thick hyphae, which spread from cell to cell by means of slen- der projections. The latter penetrate the cell wall, which becomes modified around them into a sheath-like structure, and having thus gained access to an adjoining cell con- tinue their development at its expense. The two usual forms of reproduction, by means of conidia and resting spores, are found in the genus and are of a very simple type. ‘When about to produce conidia the short thick hyphae or hyphal bodies, as they may be termed for convenience, germinate sending up asexually fructifying hyphae or conidio- phores which, after penctrating to the surface of the prothallus, become swollen at their extremities and produce ovoid conidia which are discharged into the air. After their discharge the conidia become pear-shaped, and the basal papilla of attachment to the basidium, or swollen extremity of the conidiophore, is protruded as a hyaline append- age (Nabel). The conidia germinate and spread the disease by entering other pro- thalli with which they may come in contact. The resting spores are formed within the cells of the prothallus, and result from the mere contraction of the contents of the hy- phal ‘bodies, which become surrounded by a thick wall. According to Leitgeb, this formation shows no indications whatever of a sexual origin, although his figures do not seem to preclude such a possibility in view of what is at present known of sexual proc- esses in the group. The germination and further development of these resting spores have not as yet been observed. The genus is at present represented by a single species, Completoria complens Lohde, and has been found and cultivated by Leitgeb upon pro- thalli of numerous genera and species. It is at present unknown in this country. From this comparatively simple form we may now pass to the consideration of one somewhat more complicated, which is also parasitic upon another plant, in this instance a thallo- phyte. The genus Conidiobolus was accidentally discovered by Brefeld in connection with his researches upon the Tremellini on which it is parasitic; and its discoverer, having obtained spores from cultures in which it had appeared, was enabled by cultivating them in nutritive solutions, to trace its development with the greatest completeness. The co- nidia grow readily in a decoction of horse dung, forming a mass of branched and rarely septate hyphae; which, having nearly exhausted the nutritive solution, become 138 ROLAND THAXTER ON THE broken up, through the formation of partition walls, into numerous irregularly lobulated fragments which correspond to hyphal bodies, as I use the term. These lobules appear early in the development of the hyphae, in the form of irregularly swollen projections from them; and mark the points of: origin, even at an early stage, whence the conidio-- phores are subsequently to arise. Shortly after this general disintegration of the hy- phae, single, simple conidiophores arise from each fragment, in number corresponding to the swellings above described, and produce large, ovoid conidia which are discharged in the usual way. The chief interest of the genus lies, however, in the formation of its rest- ing spores, which seem to be of sexual origin. This formation of zygospores appears as that of the conidia begins to disappear, so that both forms are at first developed side by side, while eventually the conidial formation ceases entirely,— a circumstance which seems to verify this author’s previously expressed opinion that an alternation of some regularity exists between the appearance of the two types of reproduction. In the for- mation of these zygospores, hyphae arise from swollen projections, similar to those already described as being the origin of the conidiophores, which, after a variable devel- ment, conjugate through the apposition of their swollen extremities, the contents of one extremity uniting with that of its fellow through the absorption of the intervening walls, and producing in one of them a thick-walled zygospore. Owing to a difference in size of these conjugating extremities, Brefeld was inclined to place the family among the Oomycetes; but the previous observations of Nowakowski! in Hmpusa, together with more recent studies of the family, render this improbable. In from ten days to five weeks after their formation, the resting spores were made to germinate; and sending out one or more hyphae produced usually a single conidium resembling those characteristic of the species. Of these, two are described: Conidiobolus utriculosus and C. minor, neither of which has been observed in the United States. The genus Basidiobolus, discovered by Eidam, is perhaps the most interesting of the Entomophthoreae from the unusual differentiations which accompany its asexual as well as sexual reproduction. Unlike other members of the group the species are wholly sa- prophytic, occurring naturally upon the dung of frogs and lizards after evacuation; while they may be readily cultivated in nutrient solutions similar to those employed by Brefeld in his study of Conidiobolus. According to Eidam, the fungus is present in the digest- ive tract, only in the form of spores or hyphal bodies which are dormant until they are evacuated with the faeces upon which they subsequently develop, forming large color- less hyphae with numerous cross partitions. These hyphae do not become broken up into hyphal bodies before reproduction commences, except in so far as this condition may be approached in cases similar to that figured by Eidam? where, in a concentrated nutrient solution, the segments of the hyphae become rounded; but do not, however, break apart as in Conidiobolus. In reproduction the hyphal segments may produce slender single conidiophores which, rising vertically, become greatly swollen at their ex- tremities. From the apex of this conidiophore the large conidium buds and, during its formation, the swollen extremity which bears it becomes modified by the contraction and thickening of its walls into a peculiar piece or basidiwm, which is discharged, together with the conidium, by the explosion of the slender conidiophore. 2 lhe, A. 21. ¢., Pl. 1x, fig. 10. ENTOMOPHTHOREAE OF THE UNITED STATES. 139 The formation of zygospores is also quite peculiar, and always results from the conju- gation of two adjacent cells in the same hypha, except in some instances where two co- nidia may conjugate directly. In either case conjugation is preceded by the formation of finger-like processes from either of the conjugating cells, which, arising opposite each other, are usually closely applied. Conjugation, however, does not take place by means of these processes which, at first sight, would suggest a Rhynchonema-like type; but by the absorption of the partition wall between the conjugating cells and the direct passage of the contents of one into the other. In this instance, as well as in Conidiobolus, the cell in which the zygospore is to be formed is recognizable before conjugation by its larger size. The function of the finger-like processes above mentioned seems wholly connected with the division of the nuclei which pass into them and become divided in two parts; the upper portion disappearing without becoming a new nucleus while the lower passes as a nucleus into the zygospore. ‘The zygospores are of two varieties: one, larger than the more common form, is very thick walled and covered by a peculiar brown incrustation which renders it opaque; the smaller and more usual variety was made to germinate in nutrient solutions and produced hyphae which developed the characteristic conidia of the species. Two genera remain to be mentioned: Tarichium of Cohn and Massospora Peck. The former, as has been several times pointed out by writers on Entomophthoreae, is, with- out doubt, merely the resting stage of some Hmpusa, the conidia of which are as yet unknown. In Massospora, however, which has not, I believe, been previously referred to the present family, we have a form quite peculiar, the near affinities of which cannot be determined by reason of the absence of any knowledge concerning the formation of its resting spores, or the germination of the multitudinons internal spores which char- acterize the genus. With this brief mention of the remaining genera we may now pass to a consideration of the genus HZmpusa, which, with its subdivisions, includes only entomogenous forms. I have preferred to consider these subdivisions as a whole under Hmpusa as a matter of convenience, as well as from the fact that I am not at present inclined to believe that they have more than a subgeneric value; but my reasons for this course, as well as for my use of the name Hmpusa in preference to Lntomophthora, may be better given here- after when the principal morphological differences in the species have been touched upon. Tue Genus Emrusa. Infection and production of hyphal bodies.—As has already been mentioned, infection among entomogenous Entomophthoreae results from contact with a conidial spore which, adhering to the insect host, enters its body by means of a hypha of germination. The exact method of this entrance is hardly a subject for actual observation unless, perhaps, in insects which, like many aphides, are semitransparent, and, owing to their soft integu- ment, afford an easy entrance to the hypha at almost any portion of the body. In other insects, more especially beetles, grasshoppers, ichneumons and the like, the horny integ- umgnt must diminish considerably the chances of infection; and in such cases the stig- mata or the thin membrane connecting the body segments and leg joints must be the 140 ROLAND THAXTER ON THE principal points of entrance. Infection, resulting from the ingestion of spores with the food, does not, I think, occur as is indicated by experiments with wood crickets which will be mentioned under #. Grylli; and, as a rule, the digestive tract during life does not seem to be penetrated by the fungus. After the hypha of germination has entered the body of the host, it develops with some rapidity at the expense of the softer tissues. This growth usually differs from that de- scribed in Conidiobolus from the fact that, instead of producing a branched mycelium, the hyphae multiply, not by branching and continuous growth, but by the formation of what I have previously called hyphal bodies, which consist of short, thick fragments, of very varied size and shape, that are continually reproduced by budding or division until the insect is more or less completely filled with them. In some instances these hyphal bodies have been observed as naked masses of protoplasm with an amoeboid movement, as is stated to be the case in #.colorata; but in most instances a cell wall may be dem- onstrated. In @. Grylli,at an early stage, the hyphal bodies may be seen loosely adher- ing in clusters as aresult of continued budding; but more often in this and other species they occur singly or in pairs. It is probable, however, that this mode of development is subject to considerable variation and that in some instances a mycelium may be produced directly, after the entrance of the germinating hypha. I have been unsuccessful in en- deavoring to cultivate conidia in sterilized solutions; although, by employing a drop of water in which numerous aphides had been crushed, I was enabled to obtain a fairly vigorous growth from the conidia of HL. aphidis. In this case the germinating hypha branched in all directions, forming a considerable mycelium with numerous septa; but, owing to the lack of nutriment as well as to the presence of bacteria, the hyphae soon became much attenuated and finally died. DeBary’ states that this production of a mycelium as a first result of infection occurs in EH. ovispora, H. curvispora and LF. sphaerosperma (radicans) ; but, according to Nowakowski, in his summary of the Em- pusae, the first two are not thus characterized, while my own observations of Z. sphae- rosperma do not bear out his statement that the “fungus growth” within the host is filamentous in all cases. It seems not improbable that both forms of development may occur under different conditions; but, however this may be, the termination of the first or merely vegetative condition of the fungus consists in the production of a mass of hyphal bodies which fill the host more or less completely; and in no instance, I believe, is this stage or its equivalent omitted by the direct growth of the original hyphae into conidiophores. On this assumption, in cases where a direct mycelial growth follows the entrance of the hypha of germination, if indeed such instances occur, this myce- lium must fall to pieces into hyphal bodies, before the commencement of growth the direct object of which is reproduction, in a fashion resembling that above described at a similar stage in Conidiobolus. The hyphal bodies, the production of which usually marks the end of any appropria- tion of nourishment from the host and generally occurs at about the time when the host has ceased to live, are in many cases somewhat different from those which have previ- ously characterized the fungus and often possess great regularity, both in size and shape, closely resembling spores. In #. Fresenii, for example, the original hyphal bodies 'Vergl. Morphol. d. Pilze, ete. 21. ¢., B, p. 176. ENTOMOPHTHOREAE OF THE UNITED STATES. 141 are such as are represented in figs. 106-108, while fig. 127 shows examples of those which precede the spore formation and are derived from them. In other cases, the ulti- mate hyphal bodies may be very irregular in size and shape. In all instances, they contain a highly concentrated fatty protoplasm and are capable of subsequent and often very extended development. Having reached this condition by the production of a mass of hyphal bodies, the fun- gus, under favorable conditions of temperature and moisture, may proceed at once to the completion of its development; but if these conditions are absent, a resting or chlamydosporic condition supervenes, in which the contents of each hyphal body be- come surrounded by a single wall of variable thickness according to the duration of this enforced resting stage. In this manner, the fungus may remain dormant for a considerable period until the presentation of proper conditions for further growth. How long the chlamydospores may live, I am unable to say; but I have observed their germi- nation after several weeks, and they probably retain their vitality for a much longer period, and may perhaps hibernate under certain circumstances. They form a very con- venient means for the cultivation of Empusae in water, in which they proceed at once to the formation of conidia or of resting spores. The period from first infection to the formation of chlamydospores or hyphal bodies, prior to the commencement of the repro- ductive growth, varies according to the host. In very minute and ephemeral insects, such as many gnats that are commonly attacked, the period must necessarily be short, not exceeding two or three days; but in cases where I have been enabled to observe this period, which has been unfortunately only in connection with the larger hosts, such as flies and caterpillars, it varies from six to twelve days. Germination of the hyphal bodies and chlamydospores.—Having appropriated the whole or the greater portion of the nourishment afforded by the host, the fungus is now ready to expend it in the second or reproductive stage of its growth. Under the influ- ence of a moist atmosphere and a sufficiently high temperature, the hyphal bodies “germinate” with great rapidity. The amount of moisture needed to produce this ger- mination is variable in different forms. In the common house-fly fungus (ZH. muscae), for instance, a slight change in the amount of atmospheric moisture is sufficient to pro- duce conidial formation and discharge. This is very noticeable on the seashore, where slight changes of the wind on or off the water produce a very rapid and noticeable ef- fect upon flies thus parasitized when observed in the ordinary atmosphere of the house. In other instances, more especially in those species which, unlike the house-fly fungus, are characterized in their conidial reproduction by a considerable external growth of hyphae, amuch greater degree of moisture is a necessity. Extreme cases of this kind are found in species such as H. conica or EH. sepulchralis which occur only in very moist situa- tions. In germinating, each hyphal body or chlamydospore sends out one or more hyphae which grow with great rapidity; but the manner of this germination, together with the subse- quent development of the resulting hyphae, varies considerably in different species and under different condition. In the simplest case a single hypha thus produced may grow directly to the outer air and then produce.a single conidium or set of conidia, according to the type peculiar to its conidiophores. In other cases, a single primary hypha may MEMOIRS BOSTON SOC. NAT. HIST., VOL TV. 21 142 ROLAND THAXTER ON THE branch indefinitely, each ultimate branch becoming a conidiophore similar to those of the more simple case just mentioned. This usually occurs where the conditions of growth have been very favorable and may be found side by side with the more simple form. Although the number of germinating hyphae developed from a single hyphal body is usually small, not as a rule exceeding one or two, certain instances occur in which the number is considerable. In 2. conglomerata, for example, as described by Sorokin, long hyphal bodies are found which germinate in all directions and are not unlike, in this re- spect, the hyphal bodies previously mentioned in Conidiobolus. These hyphae subse- quently branch and anastomose, forming a coherent mass which Sorokin has termed a stroma, and on which he has based a classification of H’mpusae into “Stromaticae” and * Astromaticae.” It is probable, however, that this condition is interchangeable with the more ordinary form, since in some specimens for instance of 1. apiculatus, I have found well marked stromata, while in others the direct development of the conidiophores from hyphal bodies has been distinctly traceable. The most singular modification of this kind, however, occurs in H. aphidis and virescens which are the only species thus far in which it has come under my observation. Here we have a body which appears to be of the same nature as the hyphal bodies in other species, of regular spherical form and with a highly refractive fatty contents, from which, soon after the death of the host, hyphae begin to germinate in all directions and in in- credible numbers, in a fashion that reminds one of a head of Aspergillus (figs. 239 and 261). The hyphae thus produced then branch and divide, becoming separated into a mass of irregular, short, contorted hyphae which fill the host and distend its body. This breaking to pieces of hyphae produced from hyphal bodies also.is found in species where the usual type of germination occurs: as in #. muscae, in which, just before the emer- gence of the conidiophores, the body cavity contains a mass of irregular short hyphae together with germinating hyphal bodies. Formation of conidiophores.—The germination of the hyphal bodies results either in the production of sexual or asexual resting spores (zygo- or azygospores); or of coni- diophores bearing conidia. In the latter instance hyphae, arising directly or indirectly from the hyphal bodies, grow rapidly outwards, burst through the less resistent por- tions of the host’s integument in spongy masses, in most instances of a livid white color. These masses sometimes vary to pale or bright green or dull olive, even in forms where their normal color is white; and there is considerable variation in their general appear- ance according to the species or to the conditions of their development. In some cases, they barely project beyond the body of the host and are confined to the points of emer- gence which are generally afforded by the thin intersegmental membranes through which they project in cushion-like rings as in #. muscae. In other cases, the external growth may be more extended and the masses may coalesce so as to cover the whole body with a continuous layer of conidiophores which may form a mass several times as large as the insect from which it springs. In the first and more simple case, where there is little external fungus growth, the cushion-like masses are usually formed by simple conidiophores (fig. 1) which, although each may be derived in common with many others from the same hyphal body, are yet ultimately simple, producing few or no branches outside the host’s body and giving rise ENTOMOPHTHOREAE OF THE UNITED STATES. 143 to a single conidium. In the second instance, where the external growth is greater, a different type usually occurs in the conidiophores. Not only is the external branching very considerable, but the ultimate divisions of each conidiophore are arranged in a co- rymbose or digitate fashion, as in figs. 202 and 220. This occurrence of simple and compound conidiophores in different species has led to the generic separation of the two groups; yet the distinction is by no means absolute, and intermediate forms occur, as in E. culicis, E. apiculatus and a few additional species. The converse is also true, and simple conidiophores are very commonly found in species where the type is usually com- pound. As has been already remarked, the growth of conidiophores is very rapid under favorable conditions; and an insect containing hyphal bodies, if placed in a damp, warm atmosphere, may give rise to the characteristic white masses in a few hours. Soon after the appearance of these masses the production of conidia commences, and this brings us to the next step in the conidial development. Formation and discharge of the conidia.—The terminal portion of the conidiophore, whether this be simple or one of several digitations, is termed the dasidium and is usu- ally swollen to a greater or less extent. From the apex of this basidium, which is ho- mologous with similar structures occurring among the Mucorini, the conidium, or more properly the mother cell of the conidium, is formed by budding. This bud increases at the expense of a portion of the contents of the basidium, until it has attained very nearly the normal shape and dimensions of the conidium; when it becomes separated from the basidium by a cross partition, which forms at first a horizontal plane of separa- tion between the two and is homologous with the columella of the Mucors. Within the mother cell thus formed is developed a single conidial spore, the walls of which are nor- mally in close apposition to those of its containing cell, which must thus be considered asporangium reduced to its simplest terms and modified to combine economy of material with a judicious dissemination. The resemblance to a one-spored sporangium is clearly seen in cases where, through the absorption of water by osmosis, the wall of the mother cell becomes separated from that of the conidium; a phenomenon which is very com- monly seen after, or even before, the discharge of the conidium, and is sometimes carried to such an extent that the conidium may be seen floating free in a large spherical mother cell (fig. 321). When the conidium is fully developed, and even previous to this, the contents of the spore, as well as of the basidium, begin to expand through the absorption of water. At first, as a rule, the contents of the basidium exert the greater of the two forces thus produced; perhaps owing to the fact that a more rapid absorption of water is possible through its single wall than through the double wall of the conidium. For this reason, the columella is at first forced outwards into the conidium towards which its convexity is thus turned. In some instances, especially in cases where the basidia are large and strong, this condition of things may continue until the discharge of the conidium. Such is apparently the explanation of the appearance figured by Nowakowski! which is referred to by Eidam’ as a mechanism for discharge, very different from that usually found in Empusae. I think, however, that this will prove to be only an extreme case of the nature just described. In my own experience I have observed this appearance only 11.¢. B, Pl. x1, fig. $2, ete. 2. €3,. ps, 1805 144 ROLAND THAXTER ON THE infrequently in the species referred to, H. Grylli, and in allied forms (fig. 83). It should be noted that contraction of the spore contents from any cause might also produce the same condition. In by far the majority of cases, the contents of the conidium, be- ing more dense than that of the basidium, finally exert a greater pressure and forces the columella back into the basidium, thus reversing its former position. The sum of these: opposing forces is very considerable, and as a natural result of their action a rupture of the wall ensues at the point where they are opposed, that is, in a circle round the base of the mother cell. This circle of rupture is usually very evident in discharged conidia, being indicated by a slightly ragged projection which forms a ring at the base of the papilla (fig. 85). As a result of this rupture, the conidium is discharged violently into the air, often to a considerable distance. The columella commonly remains un- broken by this discharge, although it is often greatly stretched and hangs from the ba- sidium as a tongue-like projection. In other instances it may be accidentally broken or this rupture may ke normally connected with the discharge. In the latter case, a por- tion of the protoplasm from the basidium is discharged with the spore and serves to fix it to any object with which it may come in contact. The presence or absence of a columella in different species has been made a point of structural difference to which more weight has, I think, been given than is justified by the facts. The assumption that a columella exists and is wholly or partially destroyed by the discharge of protoplasm above alluded to, in cases like #. muscae for example, | seems to me at least as warrantable as the apparently needless assumption of the absence of so characteristic a structure. Moreover, the presence of a ruptured columella is often indicated even in #. muscae by the numerous shreds that may be seen adhering to the basidium after discharge (fig. 2). The conidia and their germination.—The conidia in their normal condition are of vari- ous size and form, often varying considerably in the same species. The extremes of shape are well represented by the nearly spherical spores of 7. muscae and the slender tapering form of H. gracilis. In size they vary from about 10 in length to 754 or over. They are usually hyaline, rarely slightly colored, with a fine granular contents; or, more com- monly, contain coarsely granular protoplasm with large fat globules. In many instances, these fatty bodies are so regular in size and shape that the conidia resemble asci filled with spores; a fact which probably accounts for the statement “Flocci fertiles intus sporidiis globosis referti” in a description of what seems undoubtedly an Hmpusa, by Fries.!. The common occurrence, also, of very large single oil globules seems to have caused a similar error. The walls of the conidium are, so far as known, always smooth, without spines or similar modifications, and possess an adhesive quality which serves to attach them readily to any object, even when their discharge is unaccompanied by the mass of pro- toplasm above described. The basal portion of the spore is always more or less papil- late, the papilla being in reality that portion of the spore proper which projects from within the mother cell, from which it is distinguished by the ring of dehiscence. The conidium when discharged, if by chance it has come in contact with a suitable host, adheres to it, and sends out a hypha of germination which enters its body as pre- viously described. When placed in water the conidia give rise to one or more hyphae BPC ENTOMOPHTHOREAE OF THE UNITED STATES. 145 which branch and elongate, growing constantly more attenuated, their protoplasmic con- tents becoming separated by successive cross partitions from the empty hyphae left behind (fig. 240). It may here be mentioned that this separation by cross partitions is common in the general growth of the fungus. The hyphae produced thus from conidia have the usual characteristics: a granular protoplasmic contents which often shows a very notice- able streaming motion, and contains large oil globules and a hyaline wall. The power of germination lasts, according to Brefeld, at most only a week, or slightly more in E. radicans (EF. sphaerosperma); but in my own experience I have found the period usually much shorter than this. The period is in all probability very variable; spores that have been formed under unfavorable conditions being better able to withstand simi- lar conditions: the endurance of the spores, moreover, varies with different species. As a rule, germination takes place very soon after discharge, and if the conidium has neither fallen upon a proper host nor upon a wet surface it proceeds to form Secondary conidia.—The secondary conidium is a provision for further dissemination in case the primary spore has fallen on a substance unsuited to its proper development. The most common method of formation consists in the production of a hypha of vari- able length, which, growing vertically upwards, becomes swollen at its extremity into a basidium, and produces a conidium similar to that whence it is derived. This is dis- charged in the usual fashion, and may in turn produce tertiary, ete., conidia, in a similar way, until its vitality is exhausted or it has found a suitable lodgment. The conidio- phore formed in this process is usually simple, even if the type from which it was de- rived is digitate; yet I have seen, in a case where numerous spores had been discharged upon wet moss, that the hyphae arising from them united to form a mass of conidio- . phores of the digitate type peculiar to the species. Although the form of secondary conidium just described is most commonly found, and is apparently the normal type in all species under favorable conditions, it is subject to several interesting variations that are dependent, for the most part, upon an insufficient supply of moisture. The first of these consists in the production of a secondary conid- ium quite different from the primary, either by direct budding from it (fig. 9), or borne upon a short hypha of germination (fig. 862). These conidia are also discharged; but are apparently better suited to resist unfavorable conditions than the ordinary ones, and probably retain their power of germination much longer. The most singular modifica- tion, however, is presented by a few species allied to H’. sphaerosperma and E. Fresenii. In these forms and their allies, when the conditions of moisture are unfavorable for the ordinary process, a long, slender, capillary conidiophore is produced, on the end of which is borne a peculiar secondary conidium differing still more widely from the parent spore than in the case just mentioned. These secondary conidia (figs. 157, 191, etc.) are, with one exception, nearly almond- shaped, with noticeably thick walls, and are not discharged. Whether they ever produce tertiary spores similar to the primary ones, I have been unable to determine; but the for- mation from them of tertiary conidia similar to themselves is not uncommon. They may often be seen germinating by means of an irregular hypha which, beginning as a drop- like protuberance from the apex of the spore (fig. 119), may grow to a considerable length (fig. 122). Hidam, in his paper on Basidiobolus (Pl. 9, fig. 16), figures a mode 146 ROLAND THAXTER ON THE of germination in this genus of a related type; but as the author describes the swollen extremity of the conidiophore as a basidium, the similarity is not so striking as a com- parison of this with fig. 119, for example, in the present paper, would lead one to infer. Cystidia and rhizoids—Before leaving the conidia and conidiophores, two additional structures must be mentioned, which are of some importance morphologically. These are the so-called cystidia, or paraphyses as they have been called; and a modification of cer- tain hyphae, known as rhizoids, which serve to attach the host to the substratum on which it rests. The cystidia are usually simple hyphae, exceeding the conidiophores in size, and projecting beyond them, often to a considerable distance. In some instances they are very large (fig. 306), and readily seen with the naked eye; while in others they do not differ from the conidiophores. They are not, I think, homologues of paraphyses in other fungi, and their office is unknown; unless, perhaps, they may be considered as rhi- zoids which are functionless from their position; an explanation which seems to me very probable. A modification of the paraphyses should be mentioned which occurs in £. echinospora, a species in which, contrary to the usual type, the zygospores are very com- monly produced externally. In this case, when the sporophores have emerged from the host, certain of their number may be seen projecting beyond the rest (fig. 297). "While the process of spore formation is going on, these hyphae grow rapidly, forming ultimately a delicate network about the mass of mature zygospores. The hyphae of attachment, or rhizoids as they may be conveniently termed, consist of hyphae which, growing from the lower and outer portions of the fungus mass, attach themselves to the substratum upon which the host rests and serve to hold it firmly in position. The rhizoids may be simple or variously branched, and their termination may be variously modified into an expanded “sucker” (fig. 249). They do not, I think, enter into soft substances, and their adhesion is apparently due to the presence of a viscous secretion. They are produced with great rapidity, appearing even before the host is dead, and increasing in number with the appearance of the conidiophores. Rhizoids are con- fined to certain species, and generally accompany the digitate type of conidiophores; their presence should not, however, in my opinion, be considered of any importance as a generic distinction. It is now necessary to return once more to the condition in which we find the host filled with chlamydospores, or hyphal bodies, in order to examine the phenomena con- nected with the Formation of zygospores and azygospores.——As has been previously remarked, the germination of the hyphal bodies or chlamydospores may result in the production of conidia above described; or may lead to the formation of spores called resting spores, of a very different nature, and adapted to withstand successfully conditions that would prove fatal to the conidia ina short time. The passage to this resting condition may be accom- plished by a wholly non-sexual process, in which case the resulting spore has been termed an azygospore, or by sexual union of a type similar to that found in the Mucorini. The spores thus formed are usually of large size, spherical with one exception; with a highly refractive fatty contents; surrounded by triple walls, the outer thin and representing the wall of the mother cell, the second much thicker, and the inmost usually as thick as or thicker than the other two combined. ENTOMOPHTHOREAE OF THE UNITED STATES. 147 _The simplest process by which azygospores are formed is presented by the case in which the contents of a hyphal body become directly converted into a resting spore, usually contracting somewhat and surrounding itself with two walls, the normal third wall being represented by that of the hyphal body, within which the spore may be en- tirely free, or to which it may be closely applied. A modification of this process occurs sometimes in the case of chlamydospores, which may be transformed directly into azyg- ospores by the deposition of a third inner wall. Azygospores may also be formed in a variety of ways from hyphae of germination aris- ing from chlamydospores or hyphal bodies, or not uncommonly by direct lateral budding from them. In the first case the azygospore may be terminal (fig. 40), at the apex of a hypha of greater or less length, or may bud laterally. This process, which may be read- ily seen by cultivating chlamydospores in water, resembles at first the analogous forma- tion of conidia; the end of the hypha, however, does not produce a bud, but becomes swollen into a mother cell, which is not separated from the hypha by a cross partition, and within which the double-walled spore is formed. Still another method consists in the production of azygospores interstitially (fig. 81), which is common in certain species and leads to the occurrence of spores having very irregular shapes. These in general are the more common types of azygosporic formation, of which there are numerous slight modifications. Where true zygospores are formed, a considerable amount of variation is exhibited in the process as it occurs in different species; and al- though the sexual nature of the spore is beyond question in some instances, it is not so well marked in others, and may, as Nowakowski has suggested, represent a transitional form from the truly sexual to the wholly asexual processes. Such instances are found in H. sphaerosperma, a species in which, according to Brefeld, the production of resting spores follows the septation and anastomosis of a mass of hyphae filling the host, the spores being produced laterally from these hyphae without regard to the points of an- astomosis. In my own experience I have observed something of this sort in the legs of insects attacked by H. sphaerosperma and its near ally #. occidentalis; but in the bodies of the hosts examined, which, it should be remarked, were of a very different na- ture from those (Pieris larvae) studied by Brefeld, I found only short, contorted hyphae producing spores apparently in a wholly asexual manner (fig. 217), but associated with them, numerous instances where the budding spore was directly associated with a cross- p.utition or a slight lip-lke fold indicating the previous existence of such a partition (figs. 214-216). Whether spores thus formed should be called zygospores seems at least doubtful, and the partitions described may indicate merely the ordinary division of the hyphae. Ina very few cases, however, I have, in these two species, seen a process as well marked as that represented in fig. 197, occurring in the legs of certain hosts where, as a rule, the hyphae attain a considerable length. This certainly looks like true conjuga- tion, and may lead us to cases where the presence of a sexual union is hardly to be ques- tioned. The first instance of the latter class discovered among the Entomophthoreae is that described by Nowakowski’ in his three new species, . conica, H. ovispora and #. curvispora. In this type we have hyphae, within or without the body of the host, producing lateral outgrowths at opposite points of two different hyphae, which meet Lh Gy Aw 148 ROLAND THAXTER ON THE midway between the two conjugating cells in a fashion analogous to the similar process in Spirogyra. The intervening walls between these two gametes, or conjugating out- growths, are then absorbed and a mingling of their contents ensues. A bud then appears (fig. 322) on one or both the gametes, which increases rapidly, as a rule appropriating the entire contents of each conjugating cell to form the zygospore.