Full opinion text
FINDINGS OF FACT AND CONCLUSIONS OF LAW HOLDEN, Senior District Judge. This action was commenced by Edward and Brigitte Graham and their daughters Elizabeth, Angelica and Jean to recover for personal injuries and property damages which they allege resulted from the application of herbicides by the defendant Canadian National Railway Company (CNR) along its right of way adjacent to the Graham home place. Edward Graham deceased in January 1990. The plaintiff Brigitte has been appointed the administratrix of his estate. Mrs. Graham has been substituted as the representative of the decedent, pursuant to Rule 25(a), Fed.R.Civ.P. The complaint, as amended, is stated on various theories of legal liability; the trial proceeded on the issues of negligence and nuisance. The plaintiffs are all citizens of Vermont. The defendant CNR is a Canadian Crown Corporation with its principal office in Montreal, P.Q. It operates its rail line in Vermont. The Vermont operation of the line includes a right of way, 99 feet wide which extends 15.8 miles in the Sherbrooke Subdivision to join the Berlin Subdivision, connecting the main transcontinental Canadian line at Montreal with the CNR terminal at Portland, Maine. Jurisdiction in this court to adjudicate the present controversy is found in the provisions of the Foreign Sovereign Immunities Act, 28 U.S.C. §§ 1330(a), 1604-1607. See 14 Wright, Miller & Cooper, Federal Practice and Procedure, Jurisdiction 2d § 3662 (1985). In keeping with the jurisdictional grant, the cause was tried to the court without a jury. Bailey v. Grand Trunk Lines New England, 805 F.2d 1097, 1100 (2d Cir.), cert. denied 484 U.S. 826, 108 S.Ct. 94, 98 L.Ed.2d 54 (1987). THE FACTS Edward and Brigitte Graham purchased the premises that are at the center stage of this controversy on January 14, 1963, at a time when the rail line was in active operation. They took possession the following April. The property consists of approximately 35 acres of land with a dwelling home and outbuildings where a limited number of livestock are sheltered. The parcel is in the town of Norton, Vermont and is divided by the right-of-way. CNR traverses the Graham property south of Norton Village at mile post # 12 on the Sherbrooke Subdivision for a distance of 1,275 feet. (Ex. 052001). The Coaticook River flows west of the Graham land; state highway 114 is east of the Graham home place. APPLICATION OF HERBICIDES In the performance of routine maintenance of its track and roadbed, CNR undertook to control and destroy weeds and other vegetation on the roadbed and along its right-of-way. During the time since the Graham’s occupancy of their property, CNR has sought to eliminate the unwanted vegetation by cutting down the weeds and by application of chemical herbicides to destroy the noxious plant life. To carry out weed and brush control in the area of concern in this litigation, CNR engaged the services of independent professional chemical companies. Mowing of vegetation on and near the right-of-way was done by railroad section crews. Over the period from 1963, when the Graham family commenced the occupancy of the present home place, to 1974 the records of specific occasions of the application of toxic chemicals are indistinct and limited. William Boudle, a former employee of CNR in the capacity of section and patrol foreman during the period from 1953 until his retirement in 1987, testified chemical herbicides were sprayed on the crushed rock ballast of the roadbed for an overall width of 100 feet. Although his railroad crews did not apply the chemicals, he and other CNR employees were part of the team which guided the herbicide applicators who sprayed the line on orders from CNR offices in Montreal. The mechanical device used to apply the herbicides consisted of a track motorized vehicle equipped with booms to simultaneously spray from nozzles on both sides of the right-of-way. This method is referred to in the record as “Hy-rail” application. CNR section crews preceded the spray vehicle after proper traffic clearance was approved and signaled from the Montreal headquarters. In 1975 CNR engaged Vegetation Control Services Ltd. of Burlington, Ontario to carry out weed and brush control on the right-of-way in the Sherbrooke subdivision which included the vicinity of the Graham property. This company was granted a permit by the Vermont Department of Agriculture to spray a product referred to as Weedone 170-2.4D in the amount of one gallon per acre. The spraying is confirmed by the reports of the supervising CNR roadmaster in charge of the operations. The State of Vermont issued a similar spray permit for 1976. A copy of the application by Vegetation Control to spray CNR roadbed from Norton to the New Hampshire line bears a notation to the effect that the segment was “... eliminated 7/13/76.” (Exhibit 008004). Similar permits to spray the pertinent section of the CNR line were granted through the 1979 season to apply various herbicides, including atratol, diuron and hexazinone. Whether the herbicides were actually applied to the right-of-way adjoining the Graham farm during the years prior to 1979 is not established in the evidence. CNR admits that the companies engaged to spray its rail lines in Vermont did apply federally registered herbicides along its right-of-way in Norton, adjacent to the Graham property, in the years 1975, 1977, 1979 and 1984. The 1979 spraying of CNR’s right-of-way was conducted during July, pursuant to a permit issued by the State of Vermont at the request of the railroad for application by Vegetation Control Services Ltd. Federally registered herbicides, namely kar-mex (diuron) and velpar (hexazinone), were sprayed on the railroad ballast. Reports for this application were duly filed as required by the state agency regulations. The court finds that CNR’s right-of-way adjacent to the Graham home place was sprayed in accordance with the permit issued by the State of Vermont in 1979. OTHER CHEMICAL EVENTS The Yellow Pollen Cloud In May 1984, members of the Graham family and some of their neighbors observed a yellow cloud which appeared on their property in the area between the highway and the river, with its heaviest concentration over the railroad tracks. This observation caused Mrs. Graham to telephone the Vermont Department of Agriculture to complain that the area had been sprayed with a yellow substance which created a cloud over the property. Mrs. Graham reported her belief that the substance had been sprayed from an airplane on the adjoining property of a neighbor. Application of July 21, 1984 The main thrust of the evidence concerning CNR’s use of herbicides centers on the spraying performed July 21, 1984. This operation was undertaken pursuant to a permit issued by request of CNR’s planning engineer, Bernard Aquin, to spray from Island Pond to the New Hampshire state line. The State Agriculture Department specifically pointed out to CNR’s representatives, Aquin and the road master, Paul King, that permission to spray north of Island Pond in the Town of Norton was withheld. Other conditions were specified, including 24 hour notice to the state agency before spraying, use of special maps to observe water supply sources, waterways and streams, daily reports during the spraying operation and public notice by media advertising. One of the reasons the state agency precluded the Town of Norton was the ongoing investigation by the department which began on May 18, 1984. The permit authorized spraying of Kro-var I; its components are 40% bromacil and 40% diuron at 14.2 pounds per acre. The herbicide was applied according to instructions on the labels. CNR learned from sales representatives that Krovar had produced good results. The chemical Krovar was selected on that basis without any objective consideration as to its health effect on animal or human life. Although Aquin was on vacation, his employer took no affirmative steps to ensure that the restrictions in the state permit were observed or followed by his replacement. The restrictions imposed by the permit were not heeded by CNR, nor the company engaged to carry out the 1984 spraying contract. On July 21, 1984 the Asplundh Tree Expert Co. sprayed the CNR right-of-way in the Sherbrooke Subdivision in the area which passes through the Graham farm with the herbicides that contained atrazine, bromacil and diuron. (Ex. 044002). Although the herbicide mix was applied by Asplundh, CNR accepted full responsibility for complete compliance with the statutes and regulations of the State of Vermont. (Exhibits 069092-069095). In CNR’s request for a permit to the State Department of Agriculture, the railroad acknowledged its understanding that the authorization which followed was limited to ground application to an area 16 feet within its right-of-way. CNR concedes that the herbicide sprayed in the application on the right-of-way at the Graham property on July 21 was in violation of the terms imposed by the State of Vermont for the conduct of this operation. It was also performed without regard for adverse consequences which the operation might visit upon the health and well-being of the persons and property within reach of its exposure. Although the Hy-rail application employed by Asplundh is designed to confine the herbicide to the right-of-way, varying atmospheric conditions expand its spread by wind drift and seepage through the soil. In the spraying done on July 21st the machine continued to operate without interruption as the vehicle traversed the area of Davis Brook on the Graham land. The stream flows through a metal culvert, 45 feet in length, under the right-of-way. Ballast Maintenance On July 23, 1984, CNR undertook an operation referred to as “surfacing and realigning the ballast.” This procedure is also conducted .by a mechanical device referred to as a “ballast tamper or regulator.” The operation gathers rock ballast from its spread along the shoulders and relocates the ballast in the crib near the center of the track. The track is lifted and the ballast is tamped in place under the ties to secure the rails from movement. When this operation is conducted on a dry ballast, particles of dust and herbicide residue in the roadbed are released into the atmosphere above the adjoining land. While the ballast and surfacing operations were underway Philip Benedict, of the State Agriculture Department, went on the Graham premises in the course of continuing investigation of the owner’s prior complaints. Mr. Benedict observed the ballasting operation and a heavy cloud of dust along the railroad tracks. While the ballasting machine was operating, Benedict immediately instructed the CNR crew chief to halt the operation. He further instructed the foreman not to resume ballasting procedures until after heavy rain had soaked the area. This instruction was then communicated to Paul King, CNR’s Road-master for this segment of the line. CNR complied with the request and halted the track surfacing and lining operations at that time. CNR has not engaged in spraying or ballasting this segment of the railroad line since July 23, 1984. In July 1986 a brown substance was discovered by the Grahams on the ballast, ties and rails. About the same time unusual symptoms were observed in the farm animals; the remains of wild birds and game were found on the premises. The brown substance first seen on the railroad right-of-way was later discovered beyond the railroad right-of-way in the Graham barnyard and duck pond. There was testimony that in August of 1986 the waters of the duck pond bubbled and foamed during the heat of the day. In early summer 1987, members of the Graham family discovered a white powdery substance in and about the railroad tracks. The plaintiffs charged CNR with a “chemical spill” of caustic soda which produced burning sensations in the throats and nasal passages of members of the family exposed to the chemical. CNR acknowledged that the substance observed on the railroad tracks in July 1987 was probably caustic soda which leaked from a hopper car that was rerouted to Montreal following a derailment in New Hampshire during this time period. THE TESTING PROCESS The appearance of the yellow pollen cloud on May 18, 1984 marked the beginning of the testing procedures undertaken by the Vermont Department of Agriculture. This event also was the starting point of the investigation of the state agency on the complaint of Mrs. Edward Graham concerning the use of herbicides by CNR. Mrs. Graham reported the yellow pollen cloud and the presence of dead insects on the rails to the Vermont State Entomologist, Philip Benedict. This witness was in charge of the State’s Agricultural Department division for control of pesticides and herbicides. After receiving information regarding the presence of the unknown yellow substance, an analysis by the department apiarist confirmed that the yellow substance was pollen. (Testimony of Richard Drutchas, Ex. 045003). No pesticide was indicated in the test results. Further samples taken on July 2, 1984, again at the instance of the plaintiffs, were negative for diuron, bromacil and herbicide damage to the plant life at Graham’s place. The Plant Industry Section of the Vermont Department of Agriculture conducted a series of tests in response to Mrs. Graham’s report of the presence of unknown powder on the property on May 18, 1984. Samples were collected and analyzed for pesticides. The test results were negative. This was followed by the Department’s collection of water samples a month later from the home well, the Davis Brook, the barn well and water faucets in the Graham kitchen. The analyses revealed high levels of iron, manganese, sodium and nitrates. No organic contaminants were found. (Ex. 029001). On July 3, 1984, Richard Jensen, the Plant Industry Agent, collected two water samples for herbicide screening. The test results disclosed no presence of herbicides in the samples. Further samples were taken on July 23, 1984, following the July 21st application. Three water samples were collected from the Graham property and a vegetation sample was taken from the CNR right-of-way. The water samples were negative for pesticides; the vegetation sample from the right-of-way was positive. On this occasion the Department provided the Grahams with water sampling bottles for mailing and analysis. Samples taken by the Grahams on July 31, August 6, 13 and 27, 1984 were analyzed. No detectable levels of pesticides were found. (Exhibits 029001 and 045030-045043). At about the same time, Foster K. Palmer, D.V.M., the Graham’s veterinarian, observed an unidentified yellow substance which he and the family believed to be pollen. Later, on July 23, 1984, a whitish substance was observed by this witness on the tracks, the underlying ballast and on vegetation on both sides of the right-of-way, approximately ten feet from the- toe of the ballast. During the course of professional visits, Dr. Palmer took numerous samples, including the water supplied from separate wells to the house and barn. The samples of water from these sources were packaged, sealed and sent to the testing agency maintained by the New York State Veterinary College at Cornell University. On the basis of a preliminary report soon after the July 25th samples were submitted, Dr. Palmer advised the Graham family not to use the water from the house and barn wells on the premises for drinking, washing or cooking. He further coun-selled the Grahams to leave the property,— and that it was not safe for them or their animals. The laboratory reports, dated August 31, 1984, of samples taken from the house, barn and stream waters state “No detectable Atrazine, Diuron, Bromacil.” (Ex. 047007). Subsequent testing revealed the presence of bromacil at three parts per billion (3 ppb) (Deposition of Joseph G. Ebel, Jr., Director of Laboratory Operations for the Toxicology Department at New York State College of Veterinary Medicine, Cornell University, pages 7 and 8). Further pathological studies of specimens of animal livers, submitted by Dr. Palmer to the Cornell Laboratory for toxicology analysis, disclosed no evidence of 2, 4-D, bromacil, diuron or atrazine. (Exhibits 047008-047009). Specimens of whole blood and urine taken by Dr. Merrithew from Edward, Brigitte, Elizabeth and Angelica on September 25, 1984 and submitted to Smith Kline Clinical Laboratory, Waltham, Massachusetts, were reported negative for atrazine, bro-macil and diuron. (Exhibits 048014.5-048014.8). Early in August 1984 a claims representative of CNR, identified as L. Belanger, took samples of water from the surface wells on the Graham property. The first sample was taken from the house well; the second from the barn well. Because the limited quantity of water sampled precluded chemical analysis, germination tests were performed with cucumber seeds as a substitute method. CNR’s test report states: The ... tests indicated that the water sample identified “House Well” contains substances that suppress germination significantly and therefore gives indication of the possible presence of herbicides in this water sample. At the same time we find that the other sample identified “Barn Well” is free of similar substances. (Pltf.Ex. 102). In the spring of 1986 the plaintiff Elizabeth Graham devised her own test in cultivating bean plants grown in potting soil that were placed in separate containers. She testified that the water was supplied to the plants from discrete sources: the barn well, the old house well, the new 1985 house well and purchased bottled water. The plants supplied with water from the bottles, the barn well and the new house well grew normally; most of the plants that were watered from the old well did not germinate. The plaintiffs attached significance to the presence of a substance referred to as 2-2 methylene diphenol and 4-4 dihydroxy diphenylmethane detected in the results of the sampling taken by Dr. Palmer on July 23, 1984. The evidence presented in the testimony of Joseph G. Ebel, Jr. and Robert H. Eckerlin, D.V.M., of the staff of the laboratory toxicology program at Cornell, concluded that these compounds are not derivatives of atrazine, bromacil and diuron or by-products of the accused herbicides. Efforts to identify the substance by computer search were unavailing. There is evidence that ethyl-phenoxy-benzene is a chemical substance that may be a breakdown product of creosote used in wood preservatives. A parallel investigation was pursued by the Vermont Department of Health. The initial report of July 20, 1984 by the Director of the Division of Environmental Health stated: “We are unable to correlate the results of these analyses with any of the symptoms which you and your animals have experienced.” (Ex. 046008). However, this agency continued its investigation of the Graham complaint. (Exhibits 046001-046017). Three water samples were collected from the Graham property and a vegetation sample was taken from the CNR right-of-way. The water samples were negative for herbicides; the vegetation sample from the right-of-way tested positive. On this occasion the Agricultural Department provided the Grahams with water sampling bottles for mailing and analysis. Samples taken by the Grahams on July 31, August 6, 13 and 27, 1984 were analyzed. Diuron, bro-macil and atrazine were not detected. (Exhibits 045030, 045037 and 045042-045043). Further samples were collected by the Grahams and their neighbors in the spring of 1985. The test results derived from the series of samplings showed no detectable levels of herbicides. A summary report of the departmental activities conducted by the Vermont Department of Agriculture is set forth in a memorandum of Philip R. Benedict, Director, Plant Industry, Laboratory and Standards Division, to the Governor’s Office, dated March 28, 1985. (Exhibits 029001 and 031001). A letter dated March 22, 1985 to Mrs. Graham from Thomas J. Broido, Environmental Technician from the Vermont Department of Health, concerning the concurrent investigation conducted by that agency, disclosed that no detectable presence of bromacil, diuron or atrazine was found from the Graham house, barn, the railroad or highway culverts on the Davis Brook. The communication also advised Mrs. Graham: These results indicate that the herbicides which were applied in the area are not present in either of your water supplies at detectable levels. In addition, the volatile organic chemicals which were tested for were not present at detectable levels. The bacteria results do indicate that you may have a bacteria problem in your barn well. (Ex. 046011). The evidence presented by the plaintiffs does not establish the amount of the herbicidal chemicals that were applied by CNR on July 23, 1984 in the vicinity of the Graham premises. And there is no proof concerning the nature and extent of the exposure experienced by the individual members of the Graham family as a consequence of CNR’s weed control program during the summer of 1984, in and along the right-of-way in the Sherbrooke Subdivision at Norton. In July and August 1986, the brown substance found at the CNR farm crossing was discovered in the barnyard. Samples of mud and water were tested by the Vermont Department of Agriculture. (Ex. 045074). Bromacil, diuron and atrazine were not detected within the test limits. (Ex. 045076). A synopsis of samples taken, the times and locations, the testing laboratories and test results were compiled in a summary which was included and received in evidence as Exhibit 044001. This exhibit is annexed as an appendix to the court’s decision. The appendix is incorporated and made a part of the court's Findings of Fact. FAMILY MEDICAL HISTORY AND SYMPTOMATOLOGY The members of the Graham family were treated by Dana A. Merrithew, M.D., of the staff of the North Country Hospital and Health Center, Colebrook, New Hampshire. Dr. Merrithew recorded the medical history of the plaintiffs primarily from consultation and treatment at the Graham home and his office from 1983 to 1987. None of their ailments required inpatient hospitalization. The medical ailments experienced by the Graham family members were treated on a temporary basis as needed and were not shown by the evidence to be disabling. Edward Graham It is noted earlier in this decision that plaintiff Edward Earl Graham deceased at the age of 87 after the commencement of this action. Although his extensive medical records are included in the evidence by stipulation, no expert medical testimony was presented at trial to connect his illness to the alleged chemical events charged against CNR. The post hearing factual and legal submissions made no argument to this effect. Brigitte Graham The medical history of his widow, Brigitte, born May 20, 1931 in Germany, included respiratory problems and sporadic skin ailments manifested principally by rashes on her hands and arms. She obtained extensive medical treatment for her problems from the time the Grahams acquired the property in Norton, Vermont in 1964 to 1987. Mrs. Graham experienced intermittent manifestations of these ailments during the summer months from 1983 through 1989. Following the yellow pollen event in May 1984, her breathing difficulties lasted for periods of two hours at a time, causing sleep loss and anxiety. Skin rashes progressed to open sores in some instances. These symptoms returned following the discovery of caustic soda on the railroad bed in the summer of 1987. During the relevant period, this plaintiff complained of nausea, memory loss, impaired vision and loss of power of concentration. The medical history and diagnosis are complicated by the fact that throughout the period shown in the evidence, Brigitte Graham was an inveterate smoker. She testified she smoked at least one, and frequently more than a single pack of cigarettes daily. Elizabeth Graham Commencing in May and continuing through July 1984, Elizabeth testified she experienced many of the same maladies as her mother, although they were not reported to Dr. Merrithew until she sought medical diagnosis in 1986. At that time a skin biopsy was performed by a consulting dermatologist with inconclusive findings of chloracne and negative as to the effect of chemical exposure. (Tr. Hoffman 64-66). Angelica Graham From the early summer of 1984, Angelica suffered respiratory problems in common with her mother and sister Elizabeth. She complained of gastrointestinal ailments, skin rashes and blisters. Jean Graham, Jean testified that she suffered from the same ailments experienced by other members of the Graham family during the summer of 1984. After the July spraying by CNR in that year, Jean suffered nausea, diarrhea and severe skin eruptions on her back and face. EXPERT WITNESSES Dana Allen Merrithew, M.D., testified concerning his opinion on multiple chemical sensitivity as bearing on the causes of the plaintiffs’ medical ailments. He was the family physician, medical advisor and custodian of their medical records. Dr. Merri-thew, an internist, received his medical degree from Boston University and is certified in the specialty of internal medicine. He serves patients in the general vicinity of Colebrook, New Hampshire and Norton, Vermont. Dr. Merrithew is a general practitioner with limited training and experience in medical toxicology. He testified he did not know when he first became aware of the pathological concept of chemical sensitization. His interest was stimulated as this action advanced toward trial. The concept is of recent development according to Dr. Merrithew. Multiple chemical sensitivity afflicts certain patients who experience particular symptoms when exposed to levels of chemicals that may be below those normally expected to cause medical problems. Sensitization varies with different persons. The onset of the condition, according to the witness, is triggered by exposure to large amounts of the subject chemicals and may return by exposure to very low levels of chemicals. At a conference in his office with the plaintiffs, Brigitte, daughters Angelica and Elizabeth, and their counsel on August 17, 1985, to determine whether Mrs. Graham’s medical problems were related to pesticide exposure, Dr. Merrithew wrote in the patient’s medical record it was possible that the patient’s symptoms were related to herbicide exposure. Over strenuous objection by the defendant’s counsel, and subject to a motion to strike, Dr. Merrithew gave his opinion that Brigitte and Angelica Graham have “developed a chemical sensitization.” (Tr. Merri-thew 56). In later direct examination the witness candidly stated he didn’t have the expertise to answer which chemicals, other than the original inciting chemical, may or may not cause a recurrence of the patient’s skin rash. The witness’ response to an inquiry as to the permanency of the chemical sensitization was equivocal and speculative. The weight of the opinions expressed by Dr. Merrithew on the development of chemical sensitization and causation of the physical conditions, for which the plaintiffs seek recovery for exposure to CNR’s application of herbicides, is reduced by the shortage in the facts upon which his conclusions were based. He explained he was not familiar with the extent of exposure and the quantities of atrazine, bromacil and diuron that caused the pathological problems in the family. Except for the presence of these chemicals on July 23, 1984, he was neither informed concerning other results of the extensive testing procedures undertaken from samples from the Graham farm, nor was he aware of the test results concerning the time period involved. He was not familiar with the health advisories and the levels of danger and safety reported by the E.P.A. The witness did not relate the particular symptoms experienced by the Graham family to the specific chemicals, atra-zine, bromacil and diuron. Dr. Merrithew’s testimony provided a degree of certainty that he was “sure that the sodium hydroxide caused very specific symptoms.” (Tr. Merrithew 70). The witness went on to state that the symptoms were temporary and without lasting effect as to all members of the family. (Tr. Merrithew 84-85). At the plaintiffs’ request, Dr. Merrithew referred the Graham patients to Benjamin H. Hoffman, M.D., of Exeter, New Hampshire. Benjamin H. Hoffman, M.D., received his medical education at Mt. Sinai School of Medicine, residency at Brown University, Rhode Island Hospital in Primary Care Internal Medicine, followed by a fellowship at Yale School of Medicine in the field of Environmental Health and School of Public Health, with research in Epiderminology and Toxicology. He is presently associated with the Exeter Hospital Center for Occupational Health. Dr. Hoffman testified on the subject of the syndrome referred to as multiple chemical sensitivity. The condition is not clearly defined and is the subject of disagreement in the medical community. Patterns of the syndrome are found in persons who have been subjected to some sort of chemical exposure in the environment or industry. The reaction may be mild or acute and may differ in duration from days to weeks and months. The record establishes that Dr. Hoffman visited the premises and saw Brigitte and Elizabeth Graham on one occasion. He prescribed testing of body fluids and ordered the skin biopsy, referred to above, for Elizabeth. All pathology was negative. From the test results of samples from the Graham place, Dr. Hoffman testified atrazine, 2, 2 methylene diphenol and 4, 4 dihydroxy diphenylmethane are known skin sensitizers. Dr. Hoffman explained that these substances are contaminants present in the herbicides sprayed July 21, 1984. These chemicals have been known to produce contact dermatitis observed “in industry fairly commonly.” (Tr. Hoffman 12). The witness concluded that the symptoms experienced by Brigitte, Angelica and Elizabeth Graham, concurrently with those displayed by the animals and dead bees found on the railroad track, were caused by exposure on July 21,1984. The witness also referred to the sodium hydroxide that was discovered June 7, 1987 along the railroad right-of-way. Dr. Hoffman went on to say it was unlikely that the symptomology in the family members and animal life was caused by coliform, cigarette smoking, poor hygiene or acid rain. (Tr. Hoffman 27-32). In reaching these conclusions, the witness attached little, if any, relevance to the seventy-five negative test results reported by various agencies. The witness relied on the test performed at Cornell, which detected bromacil at 3 ppb and traces of atrazine and diuron, to conclude that this amount, although “somewhat insignificant,” suggests the presence of these chemicals elsewhere on the Graham property. The witness explained: You cannot with reasonable certainty use the water data in this case to say that this is the only exposure that these folks had. That what we have here is we know that they were exposed to these substances and, in fact, we highly suspect that they had symptoms as a result of this toxic exposure which would not easily correlate with three parts per billion. So that the suspicion is that this is a just a marker of exposure. That it is very possible that the exposure — that they did, in fact, have an effect from this exposure. (Tr. Hoffman 41-42). After serving members of the Graham family in the capacity of medical consultant over a period of one and one half years, Dr. Hoffman wrote their attorney: ... the only medical problem that is more likely than not related to their pesticide exposure is that of the chloracne that I saw on Elizabeth Graham in May of 1986. ... I feel confident that these people were exposed to a number of herbicides, including diuran [sic], bromacil, atrazine, and probably 24D, and other chlorphe-noxy herbicides. I cannot say with any certainty that any of their illnesses with the likely exception of the chloracne in Elizabeth and the possible asthma or allergic reactions in other members of the family, are definitely related to this exposure. I do feel though, they are at increased risk for developing cancer in the future. These substances (in particular 24D) have clearly been linked in a variety of different cancer studies including both in vitro, animal and human studies to the development of cancer. They have a very long latency period to the development of cancer, and therefore, an outcome cannot be measured at this time, and hopefully, they will never experience such a problem. (Pltf.Ex. 88). Ruth M. Shearer, Ph.D., testified for the plaintiffs as an expert witness in the field of toxicology. Over strenuous objection by the defendant, the court received the testimony of this witness who appeared as a consultant in the field of toxic chemicals, based on her special study of toxicology in terms of EPA regulations and advisories. Her opinion was elicited on the subject of the extrapolation of controlled experimentation with animal species to human life. Dr. Shearer received a Bachelor of Science degree in nursing from the University of Oregon in 1953, with post graduate study at the University of Washington in molecular genetics, M.S. 1966 and Ph.D., 1969. She has been consulted and has written extensively on this subject. The court finds that Dr. Shearer’s testimony served a limited, but useful, purpose in the analysis of scientific data and the nature and effect of chemical herbicides. The diagnosis of the physical and mental consequences urged by the claimants is beyond her competence. More particularly, Dr. Shearer’s effort to relate the medical history and symptoms experienced by the members of the Graham family, as factors that may contribute to multiple chemical sensitivity and her general prediction of carcinogenic effect, is without adequate foundation in medical training and experience. Frank H. Lawrence: On the subject of toxicology, the defendant presented the testimony of Frank H. Lawrence, M.D. This witness served as Chief of the Department of Emergency Medicine in Portland, Maine and Director of its Poison Control Center. In 1943, he founded and served as medical director of an environmental risk assessment firm engaged in hazard determination and exposure to chemicals and related toxic substances. His experience in this field included service of two terms of appointment as the physician member of the Maine Pesticide Control Board. The qualifications of this witness to present his opinion as a medical toxicologist are well founded in professional training and experience. Dr. Lawrence testified that the syndrome, referred to as multiple chemical sensitivity, is not totally accepted as a sound basis for reliable medical diagnosis. In this relatively recent area of toxicology, to rely on this syndrome as a proper basis for diagnosis, it is essential to have knowledge of all the chemical substances that the patient has been exposed to and the duration of that exposure in his daily life experiences. After canvassing the medical records compiled in connection with the symptoms and treatment of the members of the Graham family, Dr. Lawrence stated valid reasons for his conclusion that the plaintiffs’ ailments were not caused, within a reasonable degree of medical certainty, by the herbicides applied by CNR in 1975,1979 and 1984. More particularly, this medical witness testified that, in his opinion, there was no causal relationship between skin rashes experienced by the family members and the herbicide sprayings undertaken by the railroad. The witness explained that, in the field of toxicology, to determine the causal connection between exposure to a chemical substance and adverse health effects, the presence of the chemical must be of a sufficient amount to represent a reference dose. The mere presence of a material in the breathing zone, within reach of skin contact or subject to ingestion, is without medical significance unless the toxic substance “penetrates to the target organ of the body.” (Tr. Lawrence 17). Unless the substance is present in an amount sufficient to constitute a reference dose (RFD), no health hazard threatens the subject. When a reference dose is established and the hazardous substance contacts the human body through inhalation, ingestion or dermal exposure, causal connection is made. The dosage and the response it creates vary according to the routes of exposure from the source to the person or property affected by way of ingestion, inhalation and dermal absorption. The final medical consequence may be either adverse or without harmful effect. In rendering his medical opinion, Dr. Lawrence relied on test results reported by the Cornell laboratory. Although the tests revealed the presence of bromacil at 3 ppb with traces of atrazine and diuron, the witness assumed that all three chemicals were present in the water and ingested at 3 ppb. The witness considered the relevant factors of the toxicity of the herbicide applied by CNR, level of the dosage, duration of the exposure and the individual plaintiffs’ medical response. He concluded that the presence of the chemicals in the Graham water supply is insufficient to create a reference dose or constitute a health hazard to the Graham family. SYMPTOMS EXHIBITED IN ANIMALS AND INSECTS Members of the Graham family testified concerning a variety of circumstances and episodes which occurred in the period subsequent to the events of July 1984. In February 1985 the water in the barn well emitted a foul odor; the animals refused to drink from that source. When the hide of one of the cows was exposed to it, the evidence is to the effect that the animal’s hair fell out. Later the cow suffered abscesses on her feet and blood was detected in her milk. During much of the same time period, domestic and wild animal and bird life on the Graham property manifested unusual ailments. Wild ducks and deer were observed in various debilitated conditions. Animal life exhibited elevated body temperatures. Birds developed skin problems, irritated eyes, and their feathers molted. Dead bees were discovered by the Grahams along the rail tracks and ties during their search of the right-of-way. Foster K. Palmer, D. V.M., is a well qualified and experienced practitioner of veterinary medicine at Lyndonville, Vermont. Dr. Palmer compiled an extensive medical history of the treatment and care of the animals on the Graham place from early 1982 to the commencement of the lawsuit. In 1984, on twenty or more occasions, he was consulted by telephone and personal visits to the area where the Graham pets and livestock were sheltered and pastured. From personal knowledge of the general area of his practice and his close acquaintance with the animals on the Graham place, Dr. Palmer concluded that the symptoms exhibited by the Graham animals “seemed to be ... directly related to chemical exposure.” (Tr. Palmer 43). This opinion was modified somewhat by testimony on cross-examination to the effect that the symptomology was also consistent with that induced by fecal coliform, prevalent generally on farms in the area of his practice. In summary, Dr. Palmer testified that in his professional opinion, the underlying cause of the symptoms experienced by the animal life on the Graham farm was chronic exposure to chemicals and pesticides. No opposing professional opinion in the field of veterinary medicine was offered on this point. Nor was Dr. Palmer’s testimony in the general area of animal husbandry rebutted by the defendant. Richard Drutchas, an apiarist from the Vermont Department of Agriculture, addressed the subject of the dead bees discovered by the members of the Graham family along the CNR roadbed. The witness was called to the Graham place to investigate the “yellow cloud” in the area on May 18, 1984. He took samples of the yellow substance from puddles, the watering trough and grasses. His microscopic examination and chemical analysis confirmed the substance was pollen. No pesticides were detected in the test results. HERBICIDAL SYMPTOMS IN PLANT LIFE The testimony of the Graham family described unusual browning which appeared in the vegetation on the Graham property during the period soon after the herbicide was sprayed on July 21, 1984. During periods of heavy rain, a white foamy substance was observed cascading down the trees on the Graham property. A variety of abnormal growths, described by the Grahams to be tumors, tree cancers and fungi, was observed. Some trees exhibited defoliation at heights of 20 to 30 feet from the base. Vegetable plants in the garden plots where corn, potatoes and other vegetables were grown, appeared defoliated. Apple and wild cherry trees indicated abnormalities which the family witnesses took to be herbicidal damage. Dale R. Bergdahl, Ph.D., Professor of Forestry at the University of Vermont, presented testimony to explain these observations by the plaintiffs. Dr. Bergdahl received his doctorate in plant pathology at the University of Minnesota in 1979. The area of his expertise includes the application and pathological effect of herbicides on vegetation, as well as the diagnosis and treatment of damage to plant life from insects and disease. From scientific evidence presented by witness Bergdahl and from photograph exhibits taken by the witness and plaintiff Elizabeth Graham in the relevant time period, the court finds the killing effect from herbicides applied by CNR in the vicinity of the Graham land was limited to the tracks and ballast on the railroad right-of-way. It does not appear from evidence at hand that the herbicides, applied by the defendant during the summer of 1984, invaded or damaged the vegetation grown on the Graham land to any appreciable effect. Subsequent photographic observation of trees on the Graham land included the appearance of yellow foam draining down the yellow birch tree at the cow pen, approximately 60 feet from the railroad tracks. The witness Bergdahl explained this phenomenon occurs in birch and other trees during rainy periods. The foam is associated with wounds, such as shown in the photograph in evidence (Pltf.Ex. 79C) and is referred to as “alcoholic flux,” which is derived from saturation of the wood to display a foaming or bubbling substance. The other unusual appearances observed in 1984 by the Graham family are attributable to lichen growth, algae and fungi that sometimes develop in the natural life of growing trees. This finding is supported by the report of the State Plant Pathologist to the Deputy Commissioner of Agriculture dated August 19, 1985. PLAINTIFFS’ EXPOSURE TO HERBICIDES The extent and degree of exposure of the Graham family and their animals to herbicides applied by CNR in 1984 and the preceding years are not established in the evidence. Toxic chemicals within the classification of atrazine, bromacil and diuron are inherently mobile. The substances are toxic and, on release, may contaminate the environment. Routes of exposure to human and animal life include inhalation of dust and vapors generated by the application of these substances. Also, they are subject to ingestion by eating and drinking of contaminated liquids and produce. Herbicides may be absorbed by the skin. The effect level of exposure varies with the tolerance levels of the persons involved and the duration of the particular exposure under consideration. The environment of the Graham farm provided several routes of exposure to family and animals living on the property: by movement in the ground water through leaching or draining into family water supplies, transfer in the varying air currents of the atmosphere and dermal contact. The court finds that the highest level of herbicide chemicals on the Graham farm, as developed from the extensive laboratory procedures undertaken by the public and private agencies, was bromacil at 3 ppb detected by the Diagnostic Laboratory of the New York Veterinary College from water taken from the Graham house well on July 23, 1984. This test result is well within the margin of safety prescribed in the 1988 “Health Advisories” issued by the Office of Drinking Water, U.S. Environmental Protection Agency. The Health Advisory (HA) Program, sponsored by the Office of Drinking Water (ODW), provides information on the health effects, analytical methodology and treatment technology that would be useful in dealing with the contamination of drinking water. Health Advisories describe nonregulatory concentrations of drinking water contaminants at which adverse health effects would not be anticipated to occur over specific exposure durations. Health Advisories contain a margin of safety to protect sensitive members of the population. Health Advisories serve as informal technical guidance to assist Federal, State and local officials responsible for protecting public health when emergency spills or contamination situations occur. They are not to be construed as legally enforceable Federal standards. The HAs are subject to change as new information becomes available. (Ex. 063001). The only known presence of any of the chemicals agents detected on the Graham premises during the relevant time was bro-macil at the level of 3 ppb in the water sample from the house well on July 23, 19.84. The actual exposure of the members of the Graham household and the animal life existing on their property from the application of herbicides by CNR over the years claimed by the plaintiffs was not measured, estimated, or otherwise demonstrated in the evidence. The reports of all chemical testing and pathological examinations of specimens from the plaintiffs, their land and animals, from 1984 to 1987 were negative for herbicidal chemicals or within acceptable limits. The court finds that the application of the herbicide identified as Krovar I on July 21, 1984 by Asplundh on the right-of-way of the CNR adjacent to the Graham premises, contained bromacil and diuron. The scattered presence of toxic chemicals in various elements of the environment in unknown quantities and for unknown time periods is not necessarily dangerous to human life. (Tr. Lawrence 89). Scientific experimentation with animals and human experience, in terms of medical history, has established that persons exposed to sufficient quantities of atrazine, bromacil and diuron for durable time periods may suffer from various symptoms with adverse health effects. The probative value of the opinion of experts in toxicology on the cause of health impairment depends on identification of the substance that is the source of the danger. From this starting point, the quantity of the substance and length of time of exposure are essential to the determination of the final harm to personal and property rights at stake. The medical witnesses were in agreement that among the ingredients of the herbicide Krovar I, bromacil and diuron have the capacity to produce harmful health effects at certain levels of exposure. These ingredients are supplemented by solvents and surfactants that tend to travel in combination along routes of exposure in the air, ground and water. The main source of the plaintiffs’ evidence on causation is found in the medical opinions of Dr. Merrithew, their treating physician, and the conclusion of their consultant, Dr. Hoffman. The respective conclusions are shored up somewhat by the only veterinarian, Dr. Palmer, who appeared and testified to the effect of the unusual and untoward symptoms indicated in animal life on the Graham land. The strongest weight bearing point expressed by these experts is the concurrence of symptoms in the humans and animals that inhabited the Graham property during the time of the use of herbicides by CNR in the summer of 1984. In the opinion of these witnesses, the symptoms were related to exposure to the chemicals applied to the railroad right-of-way. Dr. Merrithew testified that Brigitte Graham’s symptoms were “related to herbicide exposure.” Furthermore, they were “quite compatible” with exposure to sodium hydroxide. (Tr. Merrithew 38). Jean’s medical problems were “compatible” with chemical exposure; and it was “quite likely” [her] symptoms were from the herbicide exposure. (Tr. Merrithew 42). As for Angelica, there was “strong likelihood” that there was some relationship in her acneform rash and the chemical exposure. (Tr. Merrithew 48). As reported earlier, the only known presence of bromacil appeared from the test results of the sample from the Graham kitchen tap taken on July 23, 1984, in the amount of 3 ppb, with traces of atrazine and diuron. Significant amounts of these chemicals were found only in the sampling taken from vegetation in the railroad right-of-way and within the target spray area. The known presence of bromacil at 3 ppb in the Graham house well, coupled with traces of atrazine and diuron in the sampling taken by Dr. Palmer on July 23, 1984, was within predictable levels of safety prescribed by EPA and the World Health Organization. The limited amount of the plaintiffs’ exposure to these substances and the short duration of the medical symptoms reported by the treating family physician and the toxicologist, Dr. Hoffman, eliminate the capability of the sprays used by CNR as the producing and efficient cause of the plaintiffs’ injuries. The same shortage of causative effect applies to the leakage of sodium hydroxide on the right-of-way in 1987. CONCLUSION The court finds that the evidence fails to establish with any reasonable medical certainty that the release of chemical herbicides during the years 1975 through 1984 and the accidental leakage of caustic soda on the right-of-way in 1987 caused the plaintiffs’ physical ailments and damage to their property as alleged in the complaint. DISCUSSION There is a shortage in the measure of proof required to establish the nature and effect of the defendant’s application of chemical herbicides in the vicinity of the Graham home place during the years 1974 through 1983. At best, only inconclusive inferences are created. The chemical events in 1984 and later have a different bearing. There is substantial evidence that CNR participated in at least three chemical events. The first of these is the application of herbicides to control the growth of noxious vegetation on July 21, 1984; the ballast and track maintenance operation two days later on July 23, 1984 and, finally, the leakage of caustic soda from the hopper car on the roadbed nearby the Graham farm on the Sherbrooke Subdivision in July 1987. STATUTE OF LIMITATIONS Before discussing the merits of the substantive claims, we turn first to the defendant’s answer that recovery is precluded on all claims that arose prior to 1984. The applicable law of Vermont is prescribed in 12 V.S.A. §§ 511 and 512. The time limitation for the commencement of actions is measured from the date the cause of action accrues. In actions for injury to the person, with narrow exceptions, the cause is “deemed to accrue as of the date of the discovery of the injury.” 12 V.S.A. § 512(4) Supp.1989. Actions for damage to personal property are governed by the similar three year time limit (§ 512(5)). See, Stevers v. E.T. & H.K. Ide Co., Inc., 148 Vt. 12, 527 A.2d 658 (1987). The court concludes that the present action was commenced within the applicable period of limitations. Plaintiffs did not discover or have valid reason to suspect that their injuries were caused by herbicides on their property until August 1, 1984. On that date the Grahams were informed by Foster Palmer, D.V.M., that the samples he had taken and sent to Cornell for testing revealed the presence of some of the ingredients of the herbicide applied on July 21, 1984. Dr. Palmer advised the Grahams at that point in time to discontinue use of the water and leave the premises. The Grahams commenced this suit on February 10, 1987, within the three year limitation prescribed for actions to recover for bodily injury and damage to personal property. LEGAL LIABILITY The theme of this extended litigation is pollution. Manifold chemical testing of public and private agencies before and after the last spray applied by CNR from samples outside the target area, detected herbicide contamination at but one site. The presence of pollutants at other test locations was negative or below detectable levels. The remaining and final question is whether the application of herbicides by the defendant CNR constituted an unlawful interference with protected personal and property interests by pollution of waters in the house well that supplied the Graham household. The interference with the use of water by pollution that affects the quality of the water may subject the wrongdoer to liability if the activity constitutes a private nuisance that is the legal cause of physical harm to the complainant. Restatement (Second) of Torts, §§ 822, 849 (1977). The Restatement (Second) of Torts expresses the law applicable to the facts of this case: § 849. In General — Pollution (1) An interference with the use of water caused by an act or conduct that is not itself a use of water but that affects the quality or quantity of the water may subject the actor to liability if the act or conduct (a) constitutes a nuisance, (b) constitutes a trespass, or (c) is negligent, reckless or abnormally dangerous with respect to the use. The polluter is subject to liability for creating a private nuisance, provided his conduct is the legal cause of the physical harm that is inflicted on the health and property interests of the complainant. That is to say, the wrongdoer has brought about an invasion of personal or property interests that the law will remedy. Although this doctrine may not have been squarely applied in the courts of Vermont, up to now it has been accepted as a sound foundation for legal liability in this circuit. Smith v. Staso Milling Co., 18 F.2d 736, 738-39 (2d Cir.1927) (L. Hand, J.). See also Reserve Mining Co. v. E.P.A., 514 F.2d 492, 29 A.L.R.Fed. 73 (8th Cir.1975), modified 529 F.2d 181 (1976). However classified in theory, the emission of toxic substances into eléments of the environment is not actionable unless the defilement is the legal cause of physical harm to the personal and property interests sought to be protected. E.g., Langle v. Kurkul, 146 Vt. 513, 517, 510 A.2d 1301 (1986); Perkins v. Vermont Hydro-Electric Corp., 106 Vt. 367, 381, 177 A. 631 (1934); In Re “Agent Orange” Product Liability Litigation, 597 F.Supp. 740, 782 (E.D.N.Y.1984); Restatement (Second) of Torts § 9, Comment a (1965). The question is whether the conduct of CNR was a substantial factor in causing the final harm to the plaintiffs and their property. This standard has been recognized as particularly appropriate in application to injury from chronic exposure to toxic chemicals where the symptoms at issue could be related to a variation or confluence of causes. Elam v. Alcolac, Inc., 765 S.W.2d 42, 174 (Mo.App.1988), cert. den., — U.S. -, 110 S.Ct. 69, 107 L.Ed.2d 36 (1989). The evidence raises no challenge to the plaintiffs’ thesis that Krovar I applied by CNR on July 21, 1984 to its right-of-way, which intersects the Graham property, is an “economic poison” as defined in the Vermont statutory law and regulatory scheme. Vermont Regulations for Control of Pesticides, 6 V.S.A. § 911(1)(A), (5)(B). The herbicide is marketed under a label which publishes precautionary instructions that it may present hazards to the environment with specific reference to workers exposed in the area to be treated “directly or through drift.” The label warns that exposure “MAY IRRITATE EYES, NOSE, THROAT AND SKIN.” The presence of known danger created the duty of reasonable care on the part of the railroad to avoid injury to the plaintiffs and their animal stock. Indifference to the consequences of dealing with a hazardous substance is lack of due care. E.g., Thompson v. Green Mountain Power Corp., 120 Vt. 478, 484, 144 A.2d 786 (1958). There is ample evidence that the railroad failed to heed the warnings on the label of the chemicals it applied to the right-of-way. CNR omitted warning those within reasonable range as to the risk of the contaminating effect of Krovar I. And finally, the application on July 21, 1984 was undertaken without proper licensing from the state agency charged with the responsibility for controlling the defendant’s use of herbicides. Thus, it is clear that the undertaking was dangerous. Risk involved in wrongful conduct is without legal effect unless the plaintiffs’ harm is proven to be a natural sequence of the defendant’s antecedent negligence. Woodcock’s Adm’r. v. Hallock, 98 Vt. 284, 290, 127 A. 380 (1925). Liability in actions founded in negligence, whether separately stated or alleged as an element of nuisance, is not established until the fact of injury is traced and connected to an act or agency that is the defendant’s responsibility. McDonnell v. Montgomery Ward & Company, 121 Vt. 221, 229, 154 A.2d 469 (1959). The presence of negligence which is not a producing cause of the damage imposes no civil liability. Perkins v. Vt. Hydro-Electric, supra, 106 Vt. at 380, 177 A. 631. Harper & James, Law of Torts, §§ 1.28, n. 16; 20.2, p. 1110 et seq. When, as here, the causal relation between negligent conduct and final harm is not obvious to the ordinary lay person, expert opinion evidence in the fields of chemistry, medical, veterinary and botanical science not only aid, but may be essential to the determination of legal cause. See, Peteet v. Dow Chemical Co., 868 F.2d 1428, 1431 (5th Cir.1989). In the developing law of toxic torts, the acceptable standard applied to expert testimony on the issue of causation is proof to a reasonable degree of scientific certainty. See, Sterling v. Velsicol Chemical Corp., 855 F.2d 1188, 1199-1201 (6th Cir.1988). This rule for weighing the competency of medical evidence on the issue of causation is in keeping with the law of Vermont. Gregory v. Hardgrove, 419 F.2d 589, 592 (2d Cir.1969); Howley v. Kantor, 105 Vt. 128, 133, 163 A. 628 (1933). In the context of this action, medical and other scientific proof are essential to eliminate the possibility of mere conjecture and speculation. The evidence provided by Dr. Hoffman,' the plaintiffs’ expert in environmental and occupational medicine, does not reach the level of reasonable medical certainty. Sterling v. Velsicol Chemical Corp., supra, 855 F.2d at 1201; See, Rheaume v. Patterson, 289 F.2d 611, 613 (2d Cir.1961). Given the presence of danger in the application of toxic chemicals, we turn to the question of whether this kind of danger culminated in actual harm to the plaintiffs and their property interests in the animal and plant life on the Graham place. The dosage of chemicals involved in the spraying that reached the plaintiffs’ home place and barnyard, as determined by reliable testing, was not shown to have been harmful to human, plant and animal life by the required measure of reasonable certainty. Sterling v. Velsicol Chemical Corp., supra, 855 F.2d at 1199-1200. The evidence bears out the point that residue detected on the vegetation growing in the target area, within the right-of-way, was insufficient to cause an adverse response by the plaintiffs or their animals in the area. Of the available and natural routes of exposure capable of conveying pollutants to the plaintiffs, their animal stock and plant life, only the water sample from the kitchen tap produced evidence of the presence of bromacil in the amount of 3 ppb. There is no substantial evidence that the plaintiffs’ exposure to this chemical ingredient of Krovar I in the limited amount detected in the water supply to the household was the legal cause of physical harm to the members of the Graham family and their property. Proximate cause is the linchpin that fastens wrongfu