STUDIES ON THE OVER-WINTERING- AND MODES OF INFECTION OF THE FIRE BLIGHT ORGANISM A Thesis Submitted in Partial Fulfillment of the Requirement for the Degree of Doctor of Philosophy in the Michigan State College of Agriculture and Applied Science, Edgar Cecil Tullls A*B*, M.A. East Lansing, Michigan, August 1927* ProQuest Number: 10008444 All rights reserved INFORMATION TO ALL USERS The quality o f this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete m anuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest. ProQuest 10008444 Published by ProQuest LLC (2016). Copyright of the Dissertation is held by the Author. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code Microform Edition © ProQuest LLC. ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 4 8 1 0 6 - 1346 Table of Contents Page I Economic Importance History of the Disease 3 Review of Literature 4 Scope of the Investigations 5 Over-wintering 6 Dissemination by Insects 15 Dissemination by Meteoric Water 27 Infection 32 Examination of Prepared Material 43 Migration of the Organism 45 Normal Anatomy of the Apple Leaf 46 Control Measures 49 Discussion 53 Summary 54 STUDIES ON THE OVER-WINTERING AND MODES OF INFECTION OF THE FIRE BLIGHT ORGANISM Economic Importance of Fire Blight in Michigan Fire blight in Michigan ranks among the more im­ portant of the diseases of the apple and pear and periodic­ ally at least causes heavy losses to the grower. These losses are not merely a reduction in the current year's crop as is the case with scab but probably more important yet is the loss of fruiting wood caused by the removal of spurs, twigs and limbs which have been invaded by the fire blight organism. This means that at least several years will elapse before a severely blighted tree will replace the lost parts with new growths and reach the stage of production possessed before it becomes diseased. In severe cases of course it be­ comes necessary to remove the tree itself. The development of the pear industry in this state is largely limited by this disease. Certain regions In the eastern and southern part of the state have been forced to cease growing this crop because of the losses caused by blight. The remains of many of these old orchards may still be seen in various parts of the state. The apple as a rule has not suffered the severe and rapid effects of the disease as has the pear but even so in cases of severe attack many apple orchards have sustained very severe losses. Apparently this loss has been more severe in the past few years. Records taken at the Huron Farms Company orchard during the summer or 1926 showed that or a block or 1306 apple trees, 166 had been so badly damaged by blight, over a period or three years, that they had to be removed. Root and crown injury in this same block was so severe that a number or trees had to be bridge grarted. Sixty-one trees had been bridge grarted in the spring or 1925. These trees were girdlea ana approach grafts were set in 1926, because it was round that this treatment was more satisfactory than attempt­ ing to clean up the cankers and make bridge grarts to the roots. One hundred firty-mine additional trees were girdled and bridge grarted in the spring or 1926. Losses in one block or 325 pears in this same orchard amounted to about 20 percent or the trees. These trees were so baaly blighted that though every attempt was made to save them, they were lost because so little was lert after the diseased limbs had been removed. A small pear orchard of 30 trees was observed during the summer or 192/, on the west side of the state, south or Sagatuck. percent of the trees were blighted. Fully 90 Twenty-five or these trees were removed the following spring. While the orchards described above represent some or the more severe infestations and losses which have been observed during the past three years, they are by no means Isolated, rare examples of the ravages of blight. sease is present and causes The di­ throughout the state a certain amount of damage in a considerable number of orchards every -o- year. Tne following data obtained, from the Plant Disease Reporter tor the years I92I-I923 inclusive, represents tne estimated percentage losses or the apple crop in Michigan, caused by fire blight and other diseases. TABUS I Percentage losses of the apple crop due to fire blight and other diseases SCAB ¥EAR BLIGHT 1921 b 2 1922 0 10 1923 b 1924 ALL OTHER DISEASES TOTAL I 8 o. 10 I I 7 b 12 2 19 I92b 8 2 3 13 AVERAGE b.6 6*6 1*4 12 .6 This table snows that losses due to fire blight are estimated by collaborators of the Plant Disease Survey to have been on an average approximately one percent lower than those due to scab for this period* History of the Disease Fire Blight has been known for about one hundred fifty years. The disease apparently is Indigenous to North America and no doubt occurred on wild hosts previous to the introduction of the apple and pear into this country. Among the wild hosts listed in North America are wild crab, hawthorne, service berry, the red berried California holly, mountain ash and the loquat. The disease has also been reported from Japan, New Zealand and Italy. It was first described in 1794 but had been observed almost fifteen years earlier by William Denning in the Hudson River Highlands. The disease spread westward and southward into the Mississippi valley and it was not until 1900 that it had reached California and In 1915 it had become a serious disease in central Washington. Nothing was known as to the cause of the disease for almost one hundred years after the disease was described. Various explanations were made In an attempt to account for its appearance* Among the causes were listed, electricity, sunscald, frozen-sap, old age, over-nourishment, under­ nourishment, fungi, Insects and other factors. Review of Literature In 1878 Burrill (b) then botanist at the University of Illinois advanced the theory that fire blight was due to a bacterial infection. This theory was substantiated by his experiments which he reported in 1881 and by the work of * Arthur (I) in 1885* Waites (2 8 ) contribution In 1898 regard­ ing the transmission of blossom infection by bees, wasps and other nectar seeking insects has been followed and supple­ mented mainly by the work of Gossard and Walton (12). Other Investigations by Jones (lb), Stewart and Leonard (2 7 ), Merrill (18) and Burrill (4 ) have established the Importance of various species of chewing and biting Insects In the dissemination of twig infection* Jones (lb), Lathrop (If) and Burrill (4) nave also made contributions concerning aphid transmission of this phase of the disease. Contribu­ tions to our knowledge of the over-wintering of the organism have been made by Sackett (23)> Whipple (29), Brooks (6) and Nixon ( 19). Miss Bachman (2) and Nixon (19) have shown the manner of intercellular migration of the organism in the in­ vasion of the host. The importance of wind blown rain has been demonstrated by Stevens (25) and his co-workers, Gossard and Walton (12) and others# Heald (15) first recognized the fact that leaf invasion produced a leaf blight. Day (9) has worked out efficient disinfectants and Reimer* has found and is testing out resistant pear stocks and our present commer­ cial varieties also in an attempt to determine which of them are most satisfactory. Scope of the Investigations The investigations reported In this paper were be­ gun in the fall of 1925 and have continued through the summer of 1928. The writer was able through arrangements with the Huron Farms Company of Ann Arbor Michigan, to spend five months during the summer of 1926 and the greater part of the * Reimer, F . C . : Blight resistance in pears and character­ istics of pear species and stocks. Sta. Bui. 214: 1-99. 1923. Ore. Agr. Exp. summer of I92y In tne Company orchards near Ann Arbor. Trips were also,taken into various parts of the southern half of the state to examine blighted orchards. Tne remainder of the work was done at the Michigan State College, Department of Botany. The problems under consideration in this paper re­ late to over-wintering of the causal organism, Bacillus amylovorus (Burr.) Trev., dissemination of the disease by aphids and by meteoric water, certain problems of Infection, and the production of twig blight. Over-wintering It Is a well established fact that the fire blight bacterium over-winters In the cankers produced on the limbs and trunks of apples, pears and quince trees. The percentage of cankers in which the causal organism remains alive through the winter in Michigan orchards has been found to vary con­ siderably in different seasons and under different conditions of vigor of the trees. Time of infection and weather con­ ditions during the late summer and fall also seem to exert an Important Influence at times. It has been commonly re­ ported and experimental evidence seem to show that the region of over-wintering is along the edge of the canker or in apparently healthy tissue just in advance of the canker edge. However cankers have been observed In which the spring exudate comes first from the more central part of the diseased area of the bark. These observations have led to an attempt to determine whether the organism may not in some cases remain alive in bark which is either dead or which con­ tains many dead cells. Two types of cankers have been observed on both apple and pear. The first may be called the determinate type of canker in which a definite margin is very evident. This is caused by the formation of a layer of cork by the host in advance of the canker edge. The second may be call­ ed the indeterminate type, for cankers which fall into this category have no definite edge and determination of the extent of the affected area is only possible by cutting through the outer bark. Cankers which have not been checked by drought or other factors are most commonly of the second type. These may result from infections of the late summer or fall. Cank­ ers formed as a result of spring and early summer Infections usually are of the determinate type* seem Some of these cankers to remain alive during the greater part of the summer and fall season and in some the bacteria seem to renew ac­ tivity in the fall after a period of slow development during the summer. In making incisions into a number of pear and apple cankers of the indeterminate type, it was found during the fall of 192b that there were certain areas outside the cambium layer in such cankers, which had a rather dark and water soaked appearance. Closer observation indicated that these areas contained bacteria in enormous numbers. Sections of this tissue gave floods of organisms similar to those ob­ tained from recently blighted twigs or bark. It was thought worth while to mark a number of cankers of the type described above and make Isolations at regular intervals through the winter and spring to watch for earliest signs of bacterial activity in the spring. Accordingly during the winter and spring of 1926 and 1927, observations were made on the over-wintering of the blight organism in pear and apple cankers. Isolations were made under microscopic observation by a modified capillary tube method, which served very satis­ factorily to obtain cultures of the organism. cases did contamination result. In only a few Bits of tissue, removed aseptically, were mounted on a slide in sterile water. Cap­ illary tubes were drawn from 2mm. soft glass tubing and fastened in a modified Barber manipulator; the capillary tube was then maneuvered until the tip was observed to be over a mass of the organisms as they exuded from the tissue. The tube was then lowered into the suspension of the organism and quickly withdrawn. The portion of the tube containing the bacteria was then broken off by means of sterile forceps and dropped into a tube of bouillon. ed within forty-eight hours* Clouding usually occurr­ These cultures were then inoc­ ulated into the host by puncture. The plants were incubated in the chambers as shown in figure 3* Isolations were made during the months of January, February, March ana April. The first exudate in the spring of 1927 was observed April 30 on pear at Ann Arbor. Cankers of apple and pear were used in making the isolations and are so designated In the table which follows. On March 8, three cultures were obtained twelve Inches back of the canker eage from an indeterminate type of apple canker. The canker was on a limb about one inch in diameter and was about three feet long. Again on April 8, three more Isolations were made from a pear canker similar to the type just described on apple. One of these cultures produced infection when Inoculated into young succulent apple shoots. Two of 26 cultures made from pear on April 29 also produced infection. The results of these inoculations, with the date of isolation of the organism, host and the region of the canker from which the isolations were made are given in the following table. TABLE 2 Results of Inoculations of cultures obtained from dormant cankers DATE OF ISOLATION NUMBER OF CULTURES OBTAINED Jan. 12 7 Febr. 6 5 Febr. 15 HOST REGION OF CANKER FROM WHICH ISOLATIONS WERE MADE RESULTS OF INOCULATIONS OF YOUNG APPLE TWIGS WITH CULTURES FROM CANKERS 5+ Pear 9 2- 4 Pear I Apple 1/4 In. back of edge I+» 3Negative Apple 1/2 In. back of edge 1+, 3- March 3* 3 Pear Center of small Negative canker around spur March 5 3 Apple 1/2 In. back of edge 3+ March 8 3 Apple 12 3+ April 5 3 Apple 1/2 In. back of edge 1+, 2- April 8 3 Pear 12 In. back of edge 1+, 2- April 9 3 2 Pear I Apple 2 In. back of edge 1+, INegative April 29 26 Pear In. back of edge Advance of discoloration 2+, 24- 4 Tne Isolations on March 3 were from material which was collected at Ann Arbor In January and February and had remained out-doors In the snow until the time of Isolation as Indicated in the table. This table shows that ten of 43 pear cankers ana that seven or 15 apple cankers contained fire blight bacteria. The percentage or active cankers may be higher earlier In tne winter but it would be necessary to work with larger numbers of cankers before any conclusions on that point might be reached. It is probable that occasionally at least, the or­ ganism may live over in the dead areas considerable distances back of the advancing edge of the canker especially in the Indeterminate type of canker. Each spring during the course of this work, blight­ ed orchards have been kept under observation, for evidences of over-wintering of tne blight organism in twigs ana blossom spu^s. Undoubtedly tne chief source of spring in­ oculum in an average season is from the cankers on limbs of more than an Inch in diameter ana from cankers on the trunk of the affected trees. There seems to be very good evidence however, as shown by Sackett (2 3 ) ana later by Brooks (o), that the blighted twigs may in some seasons serve as a limit­ ed source at least of spring Inoculum. The desirability of a full knowledge of the percentage of twigs In which the blight bacteria will live through the winter is evident since this has a direct bearing on practices of control by sanitation. Some or the evidence obtained during the course of this work Indicates that twigs may In exceptional cases be a relatively important source of spring inoculum. During the fall of 1926 a young Pippin tree was observed which had a number or ‘blighted twigs* On May 2 3 , 1927 these twigs were cut orr and examined to determine the number in which over-wintering or the organism had occurr­ ed. Invasion or the healthy tissues in advance or the canker edge had already begun, so examination was made microscopic­ ally. No Isolations were made from this material. Tne following table gives the results obtained from examination ox" this material. TABLE 3 Over-wintering of tne organism in twigs SIZK OF TWIG Vlo Vlo o/ld 6/Id 7/Id IO/lo ;> D .3 O 0 6 O * NUMBER OF TWIGS FROM WHICH FIRE BLIGHT BACTERIA WERE OBTAINED 4 I NUMBER OF TWIGS FROM WHICH NO FIRE BLIGHT BACTERIA WERE OBTAINED O IB This shows that over-wintering may occur In small twigs and it is possible that such twigs may occasionally serve as a source ox* inoculum In the spring. A large Fall Pippin in the college orchard nas been observed to have cankers on some of the large limbs for the past two springs or" 19*7 and I92d. Tne spring ox" 1927 it blossomed very heavily ana about 90 percent of tne blossoms were blighted. Five or six cankers In tne inside of the tree were observed to nave become active about tne middle of May In I92d. produced, The few blossom clusters which were became blighted. Several isolatea cases appearea in a crab tree adjacent to this tree. Numerous blossom in­ fections also appeared on another neighboring tree, the limbs of which almost interlace. The blighted blossoms appeared on those limbs closest to the Pippin tree. Twig infection was observed near an active canker in a King in which no blossom blight had appeared. This canker was located about three inches above and to one side of the infected twigs. One infection was secured on an apple seedling in the spring of 1928 from inoculum taken one-fourth of an inch back of the canker edge from apparently dead bark from a Northern Spy canker. Another case was observed in the college orchard in which over-wintering had occurred in a twig sevensixteenths of an inch. The development of the canker in the spring had cut off the terminal bud shortly after it had pro­ duced leaves. The pear orchard discussed on page 2, in which 20 percent of one block of 323 trees was lost, was due to the failure of the operator to remove a single canker on a limb about one-half inch in diameter. Observations were made in an orchard at Morrice Michigan, the first week in July 1928. trees blossom blight had occurred. In eight isolated In seven cases over­ wintering had occurred in twigs no larger than a lead pencil. It is quite obvious from the results of these ob­ servations and the observations of others that small twigs may at times serve as a primary source of Inoculum and so -14 tile necessity or the removal or these twigs should not be minimized. It was also shown that over-wintering or the blight organism might occur in dead tissue some distance back from the canker edge* Whipple (29) reports that tne pear is probably the most important host ror the over-wintering of the fire blight bacterium in Colorado, and Sackett (2 3 ) succeeded in recover­ ing the organism from 21 or blighted pear twigs. Relnking was unable to isolate tne organism from apple during the following fall, winter or spring in Wisconsin. Brooks (b) finds that over-wintering occurs in the apple in twigs as small as tnree-tenths inches in diameter. He also reports that the organism was isolated as far as one inch in advance of the canker edge in dormant cankers* Nixon has observed that pockets are formed in the tissues of the bark, along the margin of the canker. These pockets are rilled with masses or Jelly-like substance, in which the organisms are embedded. He considers that these masses or 11cysts*1 represent tne organism In its over-winter­ ing condition. The work or different investigators seems to show that these “cysts*1 may not be necessary. These so called “cysts1* have in the past been referred to by most investigators as "pockets"and were supposed to be the seats of over-wintering* Dissemination by Insects The dissemination of the fire blight organism in the orchard is one of the phases of the fire blight problem, which has been given most attention by investigators. Rain spatterings and wind blown rain are undoubtedly very import­ ant In dissemination as are insects of various kinds. Among the insects are listed, the tarnished plant bug, (Lygus pratensls Linn.), aphids, (Aphis mali Fab.), (Aphis avena Fab.),apple leaf hopper, (Empoasci mali LeBaron), curculio, (Anthonomus quadriglbbus Say.), shot hole bark boring beetle, (Scolytus rugulosus Ratzeburg), Campyloma verbasci Mey., Qrthotylus flavosparsus Sahlberg, Foeclloscytus basilis Reuter, Adelphocorls rapldus Say., Plagiognathus politus Uhler, and bees, flies and wasps. Blossom Blight Blossom Inoculations by Nectar Seeking Insects. Experimental evidence and observations on the matter of blossom infection by insects includes dissemina­ tion by nectar seeking insects and aphids. These experiments were carried on at the Huron Farms Company orchard at Ann Arbor and in the college orchard at East Lansing. During the early spring of 1926, the apple and pear orchards at Ann Arbor were very thoroughly inspected in an attempt to remove all blight cankers from the tops of the trees. The trunks and crowns of all the trees were also examined* Three cankers were found arter the trees had come into leaf. One was found in the pear block and two trunk cases were found in the apple block. Both the trunk cases in the apples were just above the surface of the ground and accordingly little transmission, if any, could have been possible from either of these sources. The canker in the pear was in the top of the tree in such a location that rain / which fell during the early part of the blossoming period, could spatter from it to blossoms in the vicinity. Because of past unproductiveness of the pear or­ chard in question, it was decided to place bees In the orchard to aid pollination* Seven hives were placed under the trees in row six figure I. The hives were placed under every third tree numbering from the bottom to the top or the chart. A bee yard was established at the same time about a hundred yards outside the main orchard, figure 2 opposite X. The pear block consists of Lawrence, Boussock and Bartlett pears. The trees represented in Figure I are large­ ly Lawrence and Boussock. The Lawrence predominating in rows 10-14, and Boussock in rows I-IO. The first pear blossoms of the various varieties opened between May 14 and May 20.Precipitation on May I4-IP-I9-2I-22. occurred The Boussocks began to blossom on May 20 and rains or .04 inches on May 21 and .4Y inches on May 22 fell during the early part of the blooming period of the Boussock variety. The canker which was found in the pears was in a Boussock tree and it is very Interesting to note that the heaviest blossom infection occurred in this variety. A survey of about half the pear block showed that many trees were badly diseased. After a thorough search only one canker which showed evidence of having been active before the main outbreak of blight took place, was found in this section of the orchard. Evidence indicates that this canker was the original source of infection which took place as shown in figure I, which gives the results of the survey. It shows the tree in which the canker was found and gives an approximate idea as to the severity of the Infection. The squares or portions which are cross hatched represent I-IO percent of Infected blossoms, the solid black represents 10-60 percent infection. By far the heaviest infection occurred in the Boussock block and isolated Boussock trees showed quite heavy infection. The whole orchard showed quite a heavy infection and it took two men about two weeks to remove the blighted parts from the affected trees over the whole 120 acres. The whole orchard was gone over twice during the spring of 1927 and any canker which had developed, after the summer cutting had been discontinued, were removed. One active canker on a Thompkins King tree was found later and one trunk canker on a pear. However neither of these trees showed any blight during the summer of 1927. May 13 and 14 blossom Inoculations were made by by spraying a water suspension of tne organism into the bloss­ oms. These inoculations were made as a rule around tne peri­ phery of the tree and not over five or six feet from the ground. The following table shows the variety inoculated, the number of blossom clusters in each variety, and the number of these inoculations which produced infection. One blossom cluster was inoculated in each of the trees shown in table 4. TABLE 4 Results of spray inoculations of blossom clusters variety: blossom Rhode Island Greening clusters INOCULATED 14 blossom clusters INFECTED 8 Wagener 17 7 Wealthy 17 4 Westfield 14 2 Grimes 7 I Duchess 3 2 North Western Greening b 3 J onathan 2 0 McIntosh 2 I Fameuse b ^j. Total 32 87 Nine of tne 32 trees In which the blossoms were infected, showed further blossom blight. From these 32 possible sources there were 110 trees infected. The writer had by examination of other apple and pear trees in the vicinity ascertained that no other sources were at hand. Some of the trees were a quarter of a mile from any other infected trees, so it seems most probable that nectar seek­ ing insects, probably bees, were agents of transmission. The weather during the blossoming period in the spring of I92Y was quite cool and a great many rainfall periods occurred during this time. The activity of bees and nectar seeking insects was no doubt below the normal and this may have accounted for the rather limited amount of spread as shown by figure 2. The largest losses occurred in the Wealthy variety which was just coming into blpssoming when these Inoculations were made and was just opposite the bee yard* It should be noted that the Jonathan, North West­ ern Greening, Rhode Island Greening, Wealthy and Grimes trees had at this time not more than half of their blossoms open. The heavy infection on the Wealthies cannot therefore be said to nave occurred because of the ract that this was the only variety still In blossom. It has been noted that years in which severe blight epidemics occur, heavy blossoming occurs also. This has been noted to be the case in a Fall Pippin tree in the college orchard. Several cankers were observed in this tree in the fall of 1920 ana again in the rail or I92Y* The tree blossomed heavily in the spring or I92Y and about 90 percent •of its blossom clusters were blighted. Severe twig blight appeared also In this tree during tne summer or I92Y* One, or tne half-dozen blossom clusters produced in tne spring 01 1928 became blighted. No twig blight appeared in this tree. It is evident tnat nectar seeking insects, during seasons of favorable weather conditions, are a considerable factor in establishing primary sources or infection in the blossoming tree. The height of their activities are no doubt best recorded in the very severe blossom epidemics which are apparent from time to time. These results corroborate the findings of Waite (28), Crandall (8 ), Jones (Io), Gossard and Walton (12) and others, that bees and other nectar seeking Insects serve as agents In the dissemination of blossom in­ fection. Infected Gossard and Walton (12) have also shown that these blossoms may serve as the primary source of infection in the tree. Hutt (14) again in 1903 proved experimentally that bees may act as agents in blossom infection. Gossard and Walton (12) show by experimental evidence and corroborate tne evidence of Waite (2 8 ) that tne bees, after visiting blossoms whicn nave become contaminated, transmit tne disease to tne next blossoms visited. It is quite evident tnat otner nec­ tar seeking insects may transmit tne disease in a like manner. Gossard and Walton may livein the noney In the (12) also snow that the organism nive seventy-two hours out they believe tnat this approaches the limit. Gtissow (13) however cites data to snow that tne organism lives as long as forty- five days when inoculated into sterile honey. Gossard and Walton (12) have evidence to show that bees may carry the organism in the pollen baskets as- well as on the mouth parts. Blossoms of peach, plum and cherry when inoculated with the organism, yielded the organism up to and including the fifth day. It is possible that blossoms so Inoculated may serve as a source of spread to apple and pear if these trees be interplanted or In adjacent plots. They report that it was observed that initial infection in a young pear orchard in which the blossoms had always been removed each year, occurr­ ed through failure to remove the blossoms. Blossom Inoculations by Aphids Some investigators have reported that aphids when hatching near the rough bark of the canker edge have crawled through the exudate and so have become contaminated with the blight organism. It is possible under such conditions that these aphids may crawl into the blossom and cause primary infection In the spring. The following experiments were carried on to de­ termine to what extent the aphids might be of importance in this respect. Twigs were cut on which large numbers of aphids were found. These twigs were then allowed to wilt so that the aphids would cease to feed and start to crawl off the twig. These aphids were removed by means of a camels hair brush to a moistened piece of bark from an active canker. They were allowed, to crawl on this bark for fifteen or twenty minutes so that their feet ana perhaps mouth parts might be­ come contaminated. Inoculation was then made by placing one aphict on the style of an apple blossom. The results of these Inoculations are given in table b« TABLE b Inoculation of apple blossoms by contaminated ahhids PLACE College Orchard BLOSSOMS INOCULATED 8b Huron Farms Co. Orchard Total BLOSSOMS INFECTED 7 BLOSSOMS NOT INFECTED 7S 89 dl ob 174 db l4o This table shows that aphids may be the cause of primary infection in the tree, but only with a certain com­ bination of weather conditions could this occur. It is probable that relatively few aphids would become so contam­ inated In any season. Twig Infection by Aphids For several years it has been generally believed by fruit growers as well as by a number of plant pathologists and horticulturists that aphids are very Important In spread­ ing blight from tree to tree. It did not seem at the beginning or the investi­ gations reported In this paper that tne evidence, presented by Jones (16), Merrill (Id) and Burrill (4) as discussed later, was extensive enough to Justify very derlnite con­ clusions regarding tne role played by these insects. The importance or tnis question rrom an economic standpoint alone would seem to justify the expenditure or considerable effort in an attempt to evaluate tne importance of these insects in the spread of fire blight* During the past three years a considerable part of the time spent in fire blight work at East Lansing and Huron Farms has been devoted to different angles of the spread of blight from blossom to twig and from twig to twig. The study of the im­ portance of aphids in this connection has necessarily been one of the major portions of this work. With this idea in mind the following experiments regarding aphid transmission were carried out during the summer of I92Y and the spring of I92b. Both green and rosy aphids were used in these inoculations. Green aphids Three aphid covered, blighting twigs or a Wagener tree were observed in the Huron Farms Company orchard, on August I, 1927. The signs or the disease had just begun to appear on these twigs even though exudation had occurred six inches back from the tip of the twig. It has been observed that under orchard conditions aphids do not remain on the diseased portions of the twig after the tissue becomes water soaked or browned. The aphids at the time or removal from the twig had ceased feeding and were crawling around on it. The aphids on healthy twigs were not crawling about. The aphids on these twigs were transferred to three succulent healthy twigs of the same variety. No disease ap- peared in these twigs. On August 9, aphids were removed from blighting twigs of Winter Banana, and were placed on lb rapidly growing, nealthy snoots of the same variety. Five aphids were put onto each twig. At the same time aphids were removed from similar blighting twigs of a Rhode Island Greening tree and transferred to bb healthy shoots of the same variety. Five aphids were placed on each shoot as in the previous experiment. The inoculated twigs remained healthy. On July 9, 192b an experiment was carried out at the nursery of the Horticulture department of the Michigan State College in a further attempt to determine how efficient aphids might be in transmitting the disease to twigs. Twigs Infested with green aphids were brought into the laboratory where they were allowed to wilt. After the a aphids had crawled off they were allowed to crawl through a water suspension of the organism. Fifteen aphids were then placed on each of 120 succulent Borsdorf twigs. were used. Forty checks These were inoculated on July II. Twenty of these were Inoculated by puncture and on 20, a water suspension was put onto the tip, leaves and the stem. These checks were in­ oculated during a rain which lasted several hours. Three weeks after the inoculations were made, two of the aphis in­ oculated shoots and nine inoculated by puncture became blight­ ed, the rest remained healthy. Rosy aphids Similar experiments were carried out in the spring of I92b in tne greenhouse and in tne nursery at the Michigan State College. The rosy aphid was used in all these experi­ ments • Twigs infested with rosy aphids were brought into the laboratory where they were allowed to wilt. After the aphids had crawled off, they were placed on blighting shoots of Transcendent Crab from which the leaves had been stripped. The blighting shoots were kept in water and the aphids trans­ ferred to them by means of a brush. These aphids were ob­ served to feed on the diseased twigs. The twigs were then placed in the foliage of potted Borsdorf seedlings in the inoculation chambers In the greenhouse to make certain that they would keep dry. The aphids crawled off the blighted twigs and colonized on 14 of the Borsdorf shoots. At the end of twenty days no blight resulted on any of these twigs. Seven days is the usual time required for symptoms of blight from puncture inoculations to become evident under the con­ ditions of these experiments. Other aphids were allowed to crawl through a sus­ pension of the organism in water. The suspension was made from macerated twigs of Borsdorf and Transcendent Crab. Five of these aphids were placed on each of 22 shoots of potted seedlings where they colonized and produced young. These plants were also put into inoculation chambers in or­ der to have the foliage remain free from water which might otherwise fall on the leaves in watering the other plants in the greenhouse. These plants were free from blight at the end of twenty days. Another similar experiment was carried on in the nursery of the Horticulture department on Borsdorf and Ben Davis seedlings. Three series of inoculations, 15, 80 and 55 twigs respectively were inoculated by placing five aphids on each twig. The aphids used had previously been allowed to crawl through a suspension of the fire blight organism. These twigs were all bagged with glassine bags to keep out rain. After three days the bags were removed and the aphids killed with Black Leaf Forty. Two days later 450 more inoculations were made in the same manner as described above except that the inoculated twigs were not bagged and the aphids were killed about twenty four hours after they were placed on the plants. One twig of the second series of bagged twigs was blighted on July 2. In all about 4500 aphids were used in these exper­ iments and about 800 twigs were inoculated. Six check shoots inoculated by puncture on June 20 became blighted at the end of ten days. It cannot be said as a result of these experi­ ments that aphids are not a factor in the dissemination of the disease to twigs. It is however quite evident that under the conditions of the experiments, very little infection was secured. Two hundred aphids were put onto one seedling to de­ termine if possible whether the number of aphids used might Influence the results of the inoculations, but this seedling also remained healthy. The results of these inoculations seem to indicate that under certain conditions at least aphids are not a considerable factor in tne dissemination or tne disease to tne twigs* In as mucn as wina blown rain is a iactor in dis­ semination of tne causal organism as was snown by Stevens (<;:?) (xossara ana Walton (12) ana otners the results obtained by Burrill are open to criticism* His experiments were carried on in tne rielu and he reported that a twig outside of his cages became infected so it is entirely possible that infect­ ion was present in the trees in the vicinity of the inoc­ ulated shoots. present. He does not state that infection was not The Infections secured might presumably have come from the outside, especially If rain had fallen, irrespective of the aphids. Merrill's experiments were made In the open and he notes that rain fell during the period that the Inocula­ tions were made. The aphids then might have been only passive carriers of the organism and not active in the pro­ duction of infection. Experimental evidence as reported by Stewart (26), Stewart and Leonard (27)* Lathrop (I?) and others leave little doubt as to the effectiveness of the other chewing and sucking Insects in the dissemination of the disease from twig to twig. Dissemination by Meteoric Water The importance of Meteoric water as a means of dis­ semination of bacterial diseases of plants has been quite thoroughly worked out in the case of some diseases. 7/hile this source of dissemination may be only relatively import­ ant, it Is probably true that in an average season it is the most consistent of all the factors which might bear on the dissemination of the blight organism. It has been noted by numerous investigators that periods of blight infection may be often correlated with rain fall periods. Because of the fact that insect transmission has been stressed in the past and because of the prevalence of these reported Insects in the orchard in any normal season, exact data on spread of fire blight by wind blown rain or rain drip will be almost impossible to secure. It is true however, that careful observations of conditions, in the orchards of Michigan under observation, may aid materially in the interpretation of conditions as they have been ob­ served to exist. Such observations were made In the early summer of 1926 at Ann Arbor and in the field at the Michigan State College in the spring of 1928. The first symptoms of blight appeared on June 14, 1926 in the pear planting of the Huron Farms Company orchard and by June 16, the symptoms had appeared in abundance in the apple trees adjacent to the pear trees. I .03 inches of rain fell on June 14, and .09 Inches fell on June 15* with .22 inches on June 21 and .07 on June 25. These rains should have been sufficient to spread the organism from the blighting blossoms to the twigs. It should be noted too that while a few rosy aphids and leaf hoppers had been observed on May 30, the orchard as a whole was almost free of these until well along into the summer when only the green aphids were present but not in sufficient numbers to Justify the use of nicotine in attempting to control them. Summer cutting of blight in the apples was begun during the last week of June and in the trees nearest the pears it was observed that twig infection often occurred be­ neath infected blossom clusters. A record was kept for two days of the instances in which these infected twigs were ob­ served to occur beneath Infected blossom clusters. On the first day 41 cases of twig blight were found beneath Infected blossom clusters. On the second day II cases of blossom infection alone were found and 38 cases of twig infection below blighted blossoms. Eleven blighted twigs were found during this period in trees which had no infected blossoms, these trees it was observed were adjacent to trees In which blossom blight occurred and it would not be impossible as shown by Faulwetter (II) that wind blown rain, or even spatters of rain from infected blossoms of neighboring trees, might have been the source of inoculum. The twigs when removed were blighted back ten to twelve inches from the tip which Is about the average dis­ tance for twigs, two weeks after inoculation, as compared to conditions found in experimental Inoculations. On July 8 , all the blighted twigs and blossoms were cut out of two rows of Wagener and Baldwin apples. Just after the two rows were completed it began to rain and sprinkled the greater of the afternoon and well Into the night. A week later on July 15, in cutting over these same two rows, 19 new twig infections and seven fruit infections were found. The latter had developed by the movement of the organisms from infected spurs dut into the fruits. These spurs no doubt served as sources of inoculum from which the twig Infection occurred, as rain fall had occurred on three successive days, on July 8-9 and 10 after the first cutting was done. The following observations were made on some small Transcendent Crab trees at the Michigan State College. Nine, seven, fifteen and sixteen blighted twigs, respectively, had been removed on May 18 from four trees. later these trees had five, seven, twigs respectively. May II to May 18. On May 25, one week two and five infected Rain had occurred during the week of No aphids or leaf hoppers were present in these trees. Other inoculations were made by fastening blighted twigs into the tops of some of the young apple and pear trees. These trees were then sprayed by an oscillating sprayer over night. The trees had also been sprayed before the inoculations were made; for one-half hour. Three twigs of one pear tree so treated became blighted after 20 days. Four twigs of one apple tree and one of another of three, similarly Inoculated trees also became blighted. From the results of observations during the past three years regarding twig infection, it seems that a close correlation exists between rainfall periods and the appearance of twig infection. Infected blossom clusters no doubt serve as the source for primary twig infection and then the first Infected twigs serve as a source of inoculum for further twig infection during the summer. It has been observed that high­ er percentages of twig infection occur In trees which have severe blossom blight. It is hardly probable that Insects might serve as carriers In this primary twig Infection as they do not attack the less succulent tissues as are found in the blighted blossoms and blossom spurs. G-ossard and Walton (12) show by experimental evi­ dence that cheese cloth bagged blossom clusters of late blossoming varieties in the vicinity of Infected trees show­ ed just as large a percentage of blight as those not bagged. So from this one might presume that insects are not the only agents of dissemination but rain too may be a considerable factor in the spread of blossom infection. Rain was also proved by them to be a most important carrier after centers of Infection had been established in the blossoming tree. From 50-90 percent of the blossom infection may be accomplish­ ed by rain. Stevens (25) and others found that wind was a fac­ tor and because they do not make any definite statement regarding rain, it has been taken by some investigators to mean that the exudate drues and becomes powdery and that this powder is blown by the wind to other trees where infection occurs. It is presumed that they refer to wind blown rain. Faulwetter (II), working on angular leaf spot of cotton, Rolfs (21), on black spot of peach, Wolf (30) on citrus canker, Rapp (22), Sackett (23), and Edgerton (10) on bean blight and Carsner (7) on angular leaf spot of cucumber all suggest that meteoric water may be an import­ ant factor In the spread of these bacterial diseases. Some of the investigators mentioned above have found rain and dew to be an important factor in the dissem­ ination and others suggest it as a possible means of spread. It seems quite probable that dissemination of diseases of this type may be accomplished by wind blown rain or dews. This is very evidently the case in the dissemination of fire blight. Infection The Relative Importance of Wounds and Natural Openings The work of various investigators on infection of the apple and pear by the blight organism has raised the question, as to the relative importance of wounds made by in­ sects, man, wind and other agencies and the natural openings in the plant tissue. It is a well known fact that usually both types of infection occur in any blighted orchard, yet which of these may be the most important is difficult to determine. Blossom Infection occurs through the nectaries of the flowers, and examination of blossoms which have been in­ oculated by spraying a water suspension of the organism into them, soon show signs of the disease. It was observed that such blossoms showed water soaked spots after three to four days, and on the second or third day following these spots had become dark brown to blackish in color. On about the tenth or twelfth day a water soaked condition could be seen on the outside of the calyx cup and drops of exudate appear­ ed several days later. The organisms are at first located among the cells of the nectary but on multiplication spread throughout the tissue of the torus and ovary and then move downward into the pedicel through the Intercellular spaces. Invasion of the fruit spur follows and many times the young fruits are in­ vaded from the spur through the pedicel. In the apple the invasion of the spur usually re­ sults in canker formation at its base. This is not always true as large cankers often result from spur Infection of this type on the more susceptible varieties. In the pear however, if conditions are at all favorable, these spur in­ fections usually result in invasion of the limb which spreads to the scaffold parts and causes large limb cankers and even occasionally results in invasion of the trunk. Twig infection as reported by numerous investigators occurs near or at the tip of the twig, at least the symptoms appear first in this region. The further invasion of the organism involves the basal part of the twig and the bark of the limb In the immediate vicinity of the base of the twig. In many cases the progress of the organism is checked at this point at least temporarily and a quite definite line appears due to the formation of a large amount of cork by the host tissue. Often however, the organism is not entire­ ly cut off from the bark tissues by this layer and further invasion follows,, TJhese periodic waves of advancement leave more or less distinct ringed markings on the bark. These concentric markings are even more marked often times in the inner than in the outer bark. Insect wounds are often found to be the seats of infection in the fruits. Several investigators have re­ corded this for curculio wounds and 12 such cases were ob­ served in the Huron Farms Company orchard in 1926. No doubt injuries caused to the fruit by worms likewise serve as seats of infection. Wounds on leaves caused by the wind whipping them against branches and fruits also may serve as sources of Inoculation. Six cases were observed in the spring of 1928 on small Transcendent Crab trees in which leaf injur­ ies of this type became Infected and twig blight resulted in all the cases. Limb injury brought about while pruning or remov­ ing blight during the growing period have become infected and Injuries brought by pickers to limbs and crotches have se served as a source of entrance of the organism. Wounds on the trunks from saw cuts, removal of suckers, and injuries from discs or harrows in cultivated orchards often times become infected and start trunk cankers. An orchard of pears near Sodus, Michigan, wnich was examined in the early summer of 1928 proved to have very nearly five percent of the 2000 trees Infected in the trunk and the majority of these cases could be traced to disc Injury on the trunk. Cultivation too close to the tree Is also very likely to wound the roots and Infection through these wounds results in the death of the tree in the majority of cases. Ten, eighteen year old pear trees at the Huron Farms Company orchard were inoculated at the crown by wound­ ing the tissues in this region with a spade, only two or three wounds were made on the crown of each tree. Two gallons of a water suspension of the organism was poured around each of these trees. through these wounds. Two of the trees became diseased Rain drip from infected twigs and blossom clusters above is no doubt the source of inoculum from which such wounds become infected. Twig Infection, on the whole, has not it seems been satisfactorily explained. Reports by various invest­ igators of severe twig infections in the absence of any of the insects which have been reported to disseminate the causal organism, raise the question as to what additional factor or factors may be concerned. Some experimental evi­ dence by Merrill (18) and Burrlll (4) seems to show that aphids may be a source of spread while experiments with about 4 5 0 0 green and rosy aphids on 800 shoots carried out at the Huron Farms Company orchard and at the Michigan State College orchard have failed to corroborate these results, only three twigs having become infected. Some inoculations made in the greenhouse during the early winter of 1927 on plants which had not been wound­ ed and on which no Insects could be found produced infection. The question arose as to how these infections had occurred. The first observed symptoms In these plants was a browning of the tips and discoloration of the tip leaves of the in­ fected shoots. From the results of inoculations and the observa­ tions of other investigators It seemed that further investi­ gation Into the matter of these leaf infections was justified. Further experiments were therefore carried out to determine the symptoms and seat of infection in these leaf lesions. Borsdorf seedling apple trees about three years old were used. Other varieties, Ben Davis, Winesap and Jonathan were used in later experiments with the same results as the Borsdorf. The potted seedlings were placed in the inocula­ tion chamber as shown in figure 3* Five or six seedlings might be inoculated in this chamber at one time. The plants were first sprayed by means of the spray attachment in the Inoculation chamber to wet the leaves thoroughly both on the upper and lower surface. The organism was then sprayed onto the under leaf surface in a water suspension by means of a DeVilbiss atomiser, or small pieces of diseased tissue about one millimeter square were placed in the droplets of water which collected on the under side of the leaves. Inoculation on the upper side of the leaves gave - 2 (- negative results except cases in which the immediate tip of the serration was included In the drop of water in which the inoculum was placed. The plants were kept in the inoculation chamber for twenty to twenty-four hours, and the leaves were kept wet by occasionally turning on the spray. The attachment was turned at an angle so that only a fine mist filled the inoc­ ulation chamber without the force of the spray touching the leaves. The plants were then removed to the bench and ex­ amined several times daily for the appearance of the symp­ toms. The period of incubation varied from two to five days depending on greenhouse conditions. Temperature seemed to affect this more than any other factor. The lesions when they first appear on the leaves are tiny black specks which are visible only by reflected light, within three or four hours these specks have Increased several times in diameter and have become water soaked in appearance. Soon they occupy the area between adjacent veinlets making an angular spot. Figure 4 represents a leaf in which the spots have reached the water soaked stage. Shortly after, the edges of the spot become browned and the progress of the organism is checked as these veinlets are reached as shown in figure 5* It is possible tnat the arresting of the organism at this point is mechanical as the intercellular spaces of the tissues surrounding the veins are very small in comparison to the size of those of the meso phyll. Within several hours upon penetration of this tissue, the water soaked condition will appear along one of the veins leading to the midrib and invasion of the shoot from the leaves follows. Figure 6 shows such invasion from two infected leaves. Invasion of the shoot has been observed to occur within 24 hours after leaf infection became visible under greenhouse conditions, and four days is the shortest time under field conditions. Early Investigators described fire blight as a di sease primary of the cambium and up to the time of Nixon's (19) publication this view was held by most plant patholo­ gists. Nixon showed by his work that the organism under optimum conditions progressed most rapidly in the outer re­ gion of the cortex and that the invasion of the cambium and medullary rays occurred last. Miss Bacnmann (2) found that occasionally the younger xylem elements were invaded also. Examination of a number of cankers in Pippin and Thompkins King trees at the Michigan State College orchard apparently bore out the findings of Nixon. The findings of the writer bore out those of Nixon relative to the nature o the matrix in which the organism is embedded. Aseptic cultures were made on August 15 and 19, 1 9 2 8 , by stripping back the bark toward the canker region. Fifty cultures were made from several twigs and small branches. regions; The material was taken from the cambium from two (I) just under the most advanced point of the canker edge, (2 ) one inch in advance of tne canker edge. One-third, of the cultures from series one and one-tenth of the cultures from series two developed bacteria which pro­ duced typical lesions when inoculated into green apples and pea r s . It seems that further investigation along this line is justified if the organism may be present as far as one inch in advance of the visible canker edge in August when the cankers were not developing rapidly. This fact may have an important bearing on the matter of over-wintering of the organism and may explain why it is possible to secure cultures an inch or so in ad­ vance of the canker edge in dormant cankers as has been shown by Brooks (6) and others. Leaf spot may also result from stomatal infections as is shown in figure 7* These spots had not developed fur­ ther, three weeks after the appearance of the symptoms. The results of the inoculations on Borsdorf and other seedlings showed no varietal differences and so no attempt will be made to differentiate between the varieties in recording the results in the following table which gives the results of these inoculations as obtained in the green­ house during the winters of 1927 and I923. TAB LE 6 Leaf inoculation of apple seedlings wltnout wounding DATE OF INOCULATION PLANTS INOCULATED PLANTS INFECTED PERCENT OF INFECTION REMARKS 1927 March 3 March 7 March 15 March 20 Oct. 17 Oct. 24 Oct. 25 Oct. 28 Oct. 29 Oct. 31 Nov. I Nov. 2 Nov. 8 N o v . 10 N ov . 13 Nov. 17 9 5 3 2 3 I 4 4 4 4 I 4 2 3 3 2 7 3 77 60 0 100 0 100 0 2 0 I 50 50 50 2 2 2 0 0 100 0 I 0 2 100 33 33 50 I I I Pure culture 1928 Febr. 7 Febr. 8 Febr. 9 Febr. II Febr. 18 March I March 7 March 8 March 10 April 18 April 24 April 30 May I May 4 May 6 May 10 Total 4 4 4 4 4 4 5 I 3 5 2 2 4 3 3 4 3 5 4 II 124 4 2 0 2 2 50 Samples taken for sectioning 50 50 25 75 2 I 4 4 4 3 7 71 100 100 100 66 33 100 80 80 75 Aphid free " " '• 63.6 " These tables show that an average of 57 percent of the plants Inoculated became infected. The largest number of infections observed on a single leaf was 28 and the average number in the different series varied quite widely with different conditions in the greenhouse at the time the plants were in the inoculation chambers. The green peach aphid was present in considerable numbers on some plants which were inoculated. However the wounds produced by these insects in feeding are very char­ acteristic and may easily be determined by use of a hand lens or microscope. Only one case of over 200 lesions ex­ amined was observed in which infection might possibly have occurred through one of these wounds. A series of inocula­ tions were also made on aphid free plants and are so noted in table 6 . Field inoculations on Transcendent Crab and pear were made as follows: the plants to be inoculated were first sprayed by means of an oscillating spray for varying lengths of time. The inoculum was put onto the young rapidly growing twigs in an aqueous suspension. for each twig. One or two drops were used The spray was then turned on again and allowed to run over night. These plants were free of aphids and leaf hoppers at the time the inoculations were made. The field inoculations have given very divergent results. One set of inoculations, in wnich four small Trans­ cendent Crab trees were used, gave 24 percent positive inocula­ tions of a total of 273 inoculated twigs. A similar set of inoculations on three trees with 89 susceptible twigs gave .07 percent of positive inoculations. Of 200 inoculations on pear twigs only five twigs became infected. Another set of inoculations of seven pear and 17 Transcendent Crab trees gave no positive results ex­ cept where the organism gained entrance through an injury on the apple leaves, ten such inoculations occurred. It is possible that the treatment before inocula­ tion was the factor influencing this divergence in the results. The plants which gave the highest percentage of infection were sprayed for five hours before inoculation. In the experiment the following day in which less than one per­ cent infection was secured the plants were sprayed for only one-half hour before being Inoculated. In the last experiment due to the fact that the trip mechanism of the sprayer failed to work, only three trees were sprayed very thoroughly before being inoculated. The results, as obtained for the pear in these in­ oculations in the field, seem to show that less twig infection occurs in the pear than in the apple. Field observations for the three years the work was in progress also corroborated this evidence. The evidence on insect transmission as presented by the various investigators aoes not seem adequate to explain twig infection under the varied conditions following which it appears. As has been pointed out twig infection frequently occurs In the absence of insects which are supposed to trans­ mit it. It is of course entirely possible and indeed it is very probable that some of these infections take place through wounds other than insect punctures or bites on tne leaves and stems. -0 0 . Xd9o. Pear blignr and its treatment. Proc. Eastern N.Y. Horr. Soc. 1897: 779-790. 1898. (29) Whipple, O.B.: Western slope fruit investigations. Pear blight. Colo. Agr. Exp. Sta. Bui. 118: 7-9* 1907. (bO) Wolr, F.A. : Citrus canker. I91o. Jour. Agric. Res. o: 09-99. SOURCE OF INOCULUM D □ □ HEAVY INFECTION ■ □ c a n □ □ LIGHT INFECTION □ □ a n a a □ □ □ □ □ □ □ a n 0 a B B S □ B B ■ H H L E □ E E a a □ E B H H I a a □ ■ aa □ □ □ □ E a b □ B h a H □ 2 □ El □ ■ □ E E □ ■ □ ma a a □ n □ □ □ □ □ □ v:\ □ El □ □ □ □ □ □ □ a a a a □ □ a a b E □ E ■ a h □ □ □ □ □ □ a □ □ a □ □ a a a a □ H p' H □ d □ □ □ ■ m E m h □ a □ n □ [ Ei H □ a □ m □ ( a □ a □ ■ E E □ □ ■ □ □ □ □ □ □ □ a e □ □ a a fi e m □ □ □ a □ b □ □ □ □ a □ a e a a a□ □ a aa □3 a4 a5 6 □ m □ □ □ □ □ □ □ ci □ 13 14 □ □ Figure I. Chart of part of the pear orchard of the Huron Farms Company at Ann Arbor Michigan. This chart shows the source of Inoculum for the other infections in the orchard during the season of 1926. It also shows the approximate severity of infection in the trees in this block. SPREAD OF BLIGHT INHURONFARMSORCHARD SEASON 1927 INOCULATED # TRANSMITTED + Figure 2. Chart of the entire orchard of the Huron Farms Company at Ann Arbor Michigan* This chart shows the spread of blight in 1927, from the trees in which blossom clusters were inoculated. Figure 3* Inoculation chamber. Green­ house inoculations were made in this type of inoculation chamber. Figure 4* Early stage of infection produced by stomatal infection. These spots were water soaked but not discolored. Figure 5* Later stage of infection than shown in figure 4. Note the dark margins of the spots. This condition follows that shown in figure 4. Figure 6* Twig blight produced by leaf infect­ ions. The third and fourth leaves were the seats of the invasion which progressed into the twig. Note the discoloration along the veins. Figure 7. Leaf spot produced by the fire blight organism. > These spots had developed no more/ three weeks after inoculation. Figure 8. Stained microtome sec­ tions of an infection which had .just come to occupy the stomatal chamber. Note the small size of the Intercellular spaces in this region and the globular shape of the cells in the mesophyll. Figure 9* A camera lucida sketch of an early stage of infection in which only the region beneath two stomata had been in­ vaded. • Figure 10. Invasion of the leaf by the zoogloeal mass. A portion of the wedge shaped invading edge is seen along the left side of the bundle This is one of the large lateral veins. Figure II. A section of the tip region of one of the serrations a short distance hack from the tip of a healthy apple leaf. The stomata, five of them* are seen along the lower surface of the leaf. The three at the left are immature. Note the small intercellular spaces above them as compared to those above the two stomata to the right. The cells of the lower epidermis are smaller, with the exception of the stomata, than those of the upper epidermis. * .raA* fev i f W &J r --.V'"' % ►i Figure 12. A series of leaves from the tip of a young Borsdorf shoot. The relative con­ ditions of maturity of these leaves are shown in table 7.