Plannan Islands Lighthouse keeper
Vienna Lighthouse in Winter 1900
▲ Cover of events — opening records
In 1900, three lighthouse keepers disappeared from a lighthouse in the Scottish archipelago of Planan. Official investigations most likely saw the possibility of being swept away by huge waves.
The mystery of the Planan Islands lies in last-minute human behavior rather than supernatural phenomena. Natural disasters are likely, but there are no complete witnesses.
Prologue: A Record-Breaking Night
When I unfolded the incident data on the desk, the first thing I noticed was not the correct answer, but the imbalance. Some scenes were too clear, and some times were entirely empty. People try to explain the whole incident by holding on to a clear scene, but I marked the empty section first. Mysteries arise because there are many clues, but sometimes the distance between them is too far.
The lighthouse was empty when the supply ship arrived at Elin Morsum. Storm damage and equipment damage were identified around the western landing site.
I didn't see this event as a work to confirm a known ending. The records were arranged in chronological order, the routes of the objects were drawn separately, and the sightings and interpretations were marked in different colors. The most famous explanation was then deliberately reviewed at the end. Because if you first read a famous hypothesis, all the clues appear to exist for that hypothesis.
▲ Reconfiguration on site — Last identified line of movement
Case Files: What I Could Believe
| Number | Verified Records | Current Reliability |
|---|---|---|
| 1 | The missing were James Dukat, Thomas Marshall and Donald MacArthur. | High |
| 2 | Lighthouse operational records and on-site damage show bad weather and high waves. | High |
| 3 | The fact that one had to remain under the regulations raised questions about the simultaneous disappearance. | the middle |
| 4 | Some of the later "flipped chairs" or grotesque diary stories don't fit the contemporary record. | the middle |
The reason why all three were outside, the final order of action, and the process of the body not being found remain.
The sentence 'unexplained' is not the same as the sentence 'supernatural' or 'someone intended'. I didn't use the blanks as evidence.
▲ Timeline — Flow before and after events
Reconfiguring Scene: The Last Definite Point
I changed the place where the incident took place to a real route rather than a stage in my head. I divided the starting point, the last check point, and the place of discovery, and calculated the time required in the meantime. Several dramatic descriptions of later generations were missing in the process. Instead, the recorded sound, weather, means of transportation, and the location of objects came forward. The fictional atmosphere arose from the texture of that very fact. There was no need to add an unidentified secret passage or an unnamed accomplice.
When I reconstructed the last scene, I did not make up the character's mind. Fear, planning, error, and fatigue can cause the same behavior. So, I wrote down my actions first and left my emotions only to the extent possible. The events became less colorful but more specific thanks to this principle.
Clue 1. The missing were James Dukat, Thomas Marshall, and Donald MacArthur.
I placed this sentence in the middle of the case record and checked the tense first. It started from who observed it, when it was written right on the spot, or whether it was added as a later memory. The longer the clues to the unsolved case get, the smoother it becomes. However, what is needed in the investigation is not the smoothness but the source. The outline of the case changed as soon as the clue was distinguished whether it was a direct observation, a trace left by an object, or someone's interpretation.
A conclusion in one direction seems quite plausible if this clue is taken aside. However, I immediately ran into the next record, "Lighthouse Operating Records and Field Damage Shows Bad Weather and High Waves." If the two do not naturally coexist in the same explanation, the explanation is still only a story. The hypothesis, in particular, fails to pass the question "How high will the waves be?" turns a blind eye to the rest of the record instead of explaining one clue.
So I made three compartments in my case notes. The first is confirmed facts, the second is reasoning that has only taken one step from facts, and the third is the possibility that you cannot choose without additional evidence. This clue is a mixture of the parts that are close to the first compartment and the parts that stay in the second compartment. I did not erase that boundary. Because leaving boundaries is the only way to correct reasoning when new records emerge later.
▲ Key Clue 1 — Traces left by records
Clue 2. Lighthouse operational records and on-site damage show bad weather and high waves.
I placed this sentence in the middle of the case record and checked the tense first. It started from who observed it, when it was written right on the spot, or whether it was added as a later memory. The longer the clues to the unsolved case get, the smoother it becomes. However, what is needed in the investigation is not the smoothness but the source. The outline of the case changed as soon as the clue was distinguished whether it was a direct observation, a trace left by an object, or someone's interpretation.
A conclusion in one direction seems very plausible if only this clue is separated. However, I immediately confronted the next record, "The fact that one person had to remain under the regulations raised the question of simultaneous disappearance." If the two do not naturally coexist in the same explanation, the explanation is still only a story. The hypothesis that fails to pass the question of "What was the urgent enough situation for the three to break the rules and go out together?" turns a blind eye to the rest of the record instead of explaining one clue.
So I made three compartments in my case notes. The first is confirmed facts, the second is reasoning that has only taken one step from facts, and the third is the possibility that you cannot choose without additional evidence. This clue is a mixture of the parts that are close to the first compartment and the parts that stay in the second compartment. I did not erase that boundary. Because leaving boundaries is the only way to correct reasoning when new records emerge later.
▲ Key clue 2 — An mismatched piece
clue 3. The fact that one person had to remain under the regulations raised the question of simultaneous disappearance.
I placed this sentence in the middle of the case record and checked the tense first. It started from who observed it, when it was written right on the spot, or whether it was added as a later memory. The longer the clues to the unsolved case get, the smoother it becomes. However, what is needed in the investigation is not the smoothness but the source. The outline of the case changed as soon as the clue was distinguished whether it was a direct observation, a trace left by an object, or someone's interpretation.
A conclusion in one direction seems very plausible if only this clue is separated. However, I immediately confronted the next record, "Some of the later 'flipped chairs' or bizarre diary stories do not fit the contemporary record." If the two do not naturally coexist in the same explanation, the explanation is still only a story. In particular, a hypothesis that fails to pass the question of "Is the absence of a body sufficiently explained in terms of ocean currents and coastal conditions?" turns a blind eye to the rest of the record instead of explaining one clue.
So I made three compartments in my case notes. The first is confirmed facts, the second is reasoning that has only taken one step from facts, and the third is the possibility that you cannot choose without additional evidence. This clue is a mixture of the parts that are close to the first compartment and the parts that stay in the second compartment. I did not erase that boundary. Because leaving boundaries is the only way to correct reasoning when new records emerge later.
▲ Key clue 3 — Evidence that shakes the hypothesis
clue 4. Some of the later 'flipped chairs' or grotesque diary stories do not fit the contemporary record.
I placed this sentence in the middle of the case record and checked the tense first. It started from who observed it, when it was written right on the spot, or whether it was added as a later memory. The longer the clues to the unsolved case get, the smoother it becomes. However, what is needed in the investigation is not the smoothness but the source. The outline of the case changed as soon as the clue was distinguished whether it was a direct observation, a trace left by an object, or someone's interpretation.
If we separate this clue, a conclusion in one direction seems very plausible. However, I immediately confronted the next record, "The missing persons were James Dukat, Thomas Marshall, and Donald MacArthur." If the two do not naturally coexist in the same explanation, the explanation is still only a story. In particular, a hypothesis that fails to pass the question "How high will the west landing site damage show?" turns a blind eye to the rest of the record instead of explaining one clue.
So I made three compartments in my case notes. The first is confirmed facts, the second is reasoning that has only taken one step from facts, and the third is the possibility that you cannot choose without additional evidence. This clue is a mixture of the parts that are close to the first compartment and the parts that stay in the second compartment. I did not erase that boundary. Because leaving boundaries is the only way to correct reasoning when new records emerge later.
Putting the three hypotheses on the same scale, the
hypothesis did not evaluate them in a nice sentence. The number of clues to be explained, the number of newly requested assumptions, and the degree to which they can withstand opposite evidence were written down, respectively. A complete explanation of one clue is not a good hypothesis if the other three are discarded.
Hypothesis A. Explanation of being swept away by exceptionally large waves while trying to protect equipment
The advantage of this explanation is that we accept with relatively few assumptions the clue that "the missing persons were James Dukat, Thomas Marshall, and Donald MacArthur." I first assumed that this hypothesis was true and re-staged the event from start to finish. I connected where the character saw what he saw, what he made, and what he moved to the next place one sentence by one. At the point of disconnection, there must have been a hidden assumption.
The biggest weakness is that it does not fully explain "one person had to remain under the regulation, which at the same time raised questions about his disappearance." If we keep adding exceptions to keep the hypothesis in place, we can explain anything, but that explanation does not predict anything. I lowered the score of the hypothesis the moment the two exceptions were exceeded. Conversely, I raised the score only when two independent clues pointed in the same direction.
In the end, this hypothesis cannot be completely discarded, but it cannot be concluded as it is. If you read an event like a novel, the blanks will naturally fill in. If you read it like an investigative record, the blanks will remain the same. I chose the latter. Even if the blanks left are uncomfortable, the inconvenience protects the extent to which the evidence can speak.
Hypothesis B. Explanation that the other two were also sacrificed while trying to rescue one person
The advantage of this explanation is that it accepts as a relatively small assumption that "Lighthouse Operating Records and Field Damage Shows Bad Weather and High Waves." I first assumed that this hypothesis was true and re-staged the event from start to finish. I connected where the character saw what he saw, what he made, and what he moved to the next place one sentence by one. At the point of disconnection, there must have been a hidden assumption.
The biggest weakness is that it does not fully explain "the later 'flipped chair' or some of the bizarre diary stories do not fit the contemporary record." If you keep adding exceptions to keep the hypothesis, you can explain anything, but such an explanation does not predict anything. I lowered the score of the hypothesis the moment the two exceptions were exceeded. Conversely, I only scored when two independent clues pointed in the same direction.
In the end, this hypothesis cannot be completely discarded, but it cannot be concluded as it is. If you read an event like a novel, the blanks will naturally fill in. If you read it like an investigative record, the blanks will remain the same. I chose the latter. Even if the blanks left are uncomfortable, the inconvenience protects the extent to which the evidence can speak.
Hypothesis C. A later explanation that there was a conflict or crime
The advantage of this explanation is that it accepts the clue that "one had to remain under the regulations" as a relatively small assumption. I first assumed that this hypothesis was true and re-staged the case from start to finish. I connected where the character saw what he saw, what he made, and what he moved to the next place one sentence by one. At the point of disconnection, there must have been a hidden assumption.
The biggest weakness is that it does not fully explain "the missing were James Dukat, Thomas Marshall, and Donald MacArthur." If you keep adding exceptions to keep the hypothesis alive, you can explain anything, but such an explanation does not predict anything. I lowered the score of the hypothesis the moment the two exceptions were exceeded. Conversely, I only scored when two independent clues pointed in the same direction.
In the end, this hypothesis cannot be completely discarded, but it cannot be concluded as it is. If you read an event like a novel, the blanks will naturally fill in. If you read it like an investigative record, the blanks will remain the same. I chose the latter. Even if the blanks left are uncomfortable, the inconvenience protects the extent to which the evidence can speak.
▲ Hypothesis Comparison — Explanatory power and gaps
The conclusions I abandoned
I was also drawn to the most famous explanation at the beginning of the investigation. It organized the case in one sentence and gave everything intention. However, when the data were compared, the explanation required a new coincidence for each blank. Then I erased the conclusion and started again. Good reasoning was more about removing unnecessary assumptions than adding anything.
The field environment strongly supports the wave hypothesis; however, the last few minutes cannot be reconstructed as there is no evidence of direct recording of the positions and judgments of the three.
In particular, the details repeated on the Internet and later testimonies were separated from the original record. Just because the same sentence appears in multiple places does not mean that there are multiple independent sources. The possibility that one wrong sentence was duplicated was first confirmed. The size of the case decreased after this process, but the weight of the remaining clues increased.
▲ Counterfeit Checks — The Conclusion I Abandoned
Final Reason: The Results I Reached
The mystery of the Planan Islands lies in last-minute human behavior rather than supernatural phenomena. Natural disasters are likely, but there are no complete witnesses.
This conclusion is not a complete final decision. It is a direction that requires the least assumptions in the currently disclosed records and explains the most clues at the same time. It can be corrected when new evidence comes out. However, it is also clear that the conclusion changes when it comes out. Reliable original records, verifiable legal and scientific data, and independent evidence that directly connects objects and people are needed.
I don't write that I closed the case. Instead, write how far you can close it. That boundary is the result of this investigation. The fun of the mystery is not that anything is possible. It is in finding out what can be thrown away among the possible explanations.
▲ Final Judgment — Description that requires the least assumption
Additional Investigation Records 1. Redraw the Timeline
At dawn, the tables were opened again. This time, he only covered up the conclusion and left a record saying, "The missing persons were James Dukat, Thomas Marshall, and Donald MacArthur." Even events that he thought he had already known will take on a completely different shape when the conclusion is erased. I connected the place and time in a line, and wrote the minimum time required to travel. The spaces between the records were marked with blank squares instead of filled in by imagination. The larger the square, the less reliable the hypothesis was to explain the section.
On top of it, I posted a hypothesis, "an explanation that I was swept away by an exceptionally large wave while trying to protect the equipment." At first, everything seemed to fit. However, when the question, "How high would the damage to the western landing site show the height of the wave?" was substituted, the weak part of the explanation was revealed. Hypotheses must endure questions. If you keep calling new people or coincidences to avoid questions, from that moment on, it becomes creation, not reasoning. I used my sense of creation only in the atmosphere of the sentence and did not put it in the place of fact.
The result I got from this section is not the name of the criminal or the complete decryption. Instead, it became clear which explanations required fewer assumptions and what would be reversed if new findings were made. Even if the case is not completed, it is possible to narrow the scope of the investigation. This record is intended to leave that narrow boundary.
Further investigation records 2. Missing links
At dawn, the event table was reopened. This time, he only covered up the conclusion and left a record saying, "Lighthouse operation records and on-site damage show bad weather and high waves." Even incidents that I thought I already knew will take on a completely different shape if I erase the conclusion. I connected the place and time in a single line and wrote the minimum time required to move. The spaces between the records were marked with blank squares instead of filled with imagination. The larger the square, the less reliable the hypothesis explaining the section was.
On top of that, I posted a hypothesis that "an explanation was that the other two were also sacrificed while trying to rescue one person." At first, everything seemed to fit together. However, when the question, "What was the urgent enough situation for the three to break the rules and go out together?" was substituted, the weak part of the explanation was revealed. Hypotheses must endure questions. If you keep calling new people or coincidences to avoid questions, from that moment on, it becomes creation, not reasoning. I used my sense of creation only in the atmosphere of the sentence and did not put it in the place of fact.
The result I got from this section is not the name of the criminal or the complete decryption. Instead, it became clear which explanations required fewer assumptions and what would be reversed if new findings were made. Even if the case is not completed, it is possible to narrow the scope of the investigation. This record is intended to leave that narrow boundary.
Additional Investigation Records 3. Eyewitness Distance
At dawn, the event table was reopened. This time, he only left a record of "One person had to remain under the regulations, raising the question of simultaneous disappearance." Even incidents that he thought he had already known will take on a completely different shape when the conclusion is deleted. I connected the place and time in a line and wrote the minimum time required to move. The spaces between the records were marked with blank squares instead of filled with imagination. The larger the square, the less reliable the hypothesis explaining the section was.
On top of that, I put a hypothesis that "a later explanation that there was a conflict or crime." At first, everything seemed to fit. However, when the question, "Is the absence of a body sufficiently explained by sea currents and coastal conditions?" was substituted, the weak part of the explanation was revealed. Hypotheses must endure questions. If you keep calling new people or coincidences to avoid questions, from that moment on, it becomes creation, not reasoning. I used my sense of creation only in the atmosphere of the sentence and did not put it in the place of fact.
The result I got from this section is not the name of the criminal or the complete decryption. Instead, it became clear which explanations required fewer assumptions and what would be reversed if new findings were made. Even if the case is not completed, it is possible to narrow the scope of the investigation. This record is intended to leave that narrow boundary.
Further investigation record 4. The path through which the item passed
At dawn, the event chart was opened again. This time, the conclusion was covered up and only a record saying, "Some of the later 'flipped chairs' or bizarre diary stories do not fit the contemporary record." Events that I thought I already knew will take on a completely different shape if I erase the conclusion. I connected the place and time in a line and wrote the minimum time required to move. The spaces between the records were marked with blank squares instead of filled with imagination. The larger the square, the less reliable the hypothesis was to explain the section.
On top of it, I posted a hypothesis, "an explanation that I was swept away by an exceptionally large wave while trying to protect the equipment." At first, everything seemed to fit. However, when the question, "How high would the damage to the western landing site show the height of the wave?" was substituted, the weak part of the explanation was revealed. Hypotheses must endure questions. If you keep calling new people or coincidences to avoid questions, from that moment on, it becomes creation, not reasoning. I used my sense of creation only in the atmosphere of the sentence and did not put it in the place of fact.
The result I got from this section is not the name of the criminal or the complete decryption. Instead, it became clear which explanations required fewer assumptions and what would be reversed if new findings were made. Even if the case is not completed, it is possible to narrow the scope of the investigation. This record is intended to leave that narrow boundary.
Additional Investigation Records 5. The Temptation of False Clues
At dawn, the tables were opened again. This time, he only covered up the conclusion and left a record saying, "The missing persons were James Dukat, Thomas Marshall, and Donald MacArthur." Even events that he thought he had already known will take on a completely different shape when the conclusion is erased. I connected the place and time in a line, and wrote the minimum time required to travel. The spaces between the records were marked with blank squares instead of filled in by imagination. The larger the square, the less reliable the hypothesis was to explain the section.
On top of that, I posted a hypothesis that "an explanation was that the other two were also sacrificed while trying to rescue one person." At first, everything seemed to fit together. However, when the question, "What was the urgent enough situation for the three to break the rules and go out together?" was substituted, the weak part of the explanation was revealed. Hypotheses must endure questions. If you keep calling new people or coincidences to avoid questions, from that moment on, it becomes creation, not reasoning. I used my sense of creation only in the atmosphere of the sentence and did not put it in the place of fact.
The result I got from this section is not the name of the criminal or the complete decryption. Instead, it became clear which explanations required fewer assumptions and what would be reversed if new findings were made. Even if the case is not completed, it is possible to narrow the scope of the investigation. This record is intended to leave that narrow boundary.
Additional Investigation Records 6. Reconstruction of the Last Scene
At dawn, the event table was reopened. This time, he only covered up the conclusion and left a record saying, "Lighthouse operation records and on-site damage show bad weather and high waves." Even incidents that I thought I already knew will take on a completely different shape if I erase the conclusion. I connected the place and time in a single line and wrote the minimum time required to move. The spaces between the records were marked with blank squares instead of filled with imagination. The larger the square, the less reliable the hypothesis explaining the section was.
On top of that, I put a hypothesis that "a later explanation that there was a conflict or crime." At first, everything seemed to fit. However, when the question, "Is the absence of a body sufficiently explained by sea currents and coastal conditions?" was substituted, the weak part of the explanation was revealed. Hypotheses must endure questions. If you keep calling new people or coincidences to avoid questions, from that moment on, it becomes creation, not reasoning. I used my sense of creation only in the atmosphere of the sentence and did not put it in the place of fact.
The result I got from this section is not the name of the criminal or the complete decryption. Instead, it became clear which explanations required fewer assumptions and what would be reversed if new findings were made. Even if the case is not completed, it is possible to narrow the scope of the investigation. This record is intended to leave that narrow boundary.
▲ End of investigative record — questions still to be asked
· Northern Lighthouse Board — Flannan Isles
· Historic Environment Scotland — Eilean Mòr Lighthouse
※ This article is a commentary in the form of personal investigation records based on public records and reliable data. We have distinguished between facts and reasoning, and when a new official announcement is made, the judgment is updated.
Image Source and Edit
We used Wikipedia Commons public license data cut, ratio adjustment, and color correction to fit the body flow. Circles with generated images are not included in the list below.
- Flannan isles.jpg — Rossser1954 - Roger Griffith, Public domain
- December Storm IMG 0182 (16000254795).jpg — Ronnie Robertson, CC BY-SA 2.0
- Ordnance Survey One-Inch Sheet 12 Isle of Lewis and North Harris, Published 1959.jpg — Director General of the Ordnance Survey, Southampton, UK, Public domain
- Buckie Lighthouse Highlands Scotland (49224497737).jpg — m.shattock, CC BY-SA 2.0
- Admiralty Chart No 3331 Flannan Isles, Published 1903.jpg — United Kingdom Hydrographic Office, Public domain
- The storm before the calm - geograph.org.uk - 3377350.jpg — Ian Paterson, CC BY-SA 2.0
'Unsolved case' 카테고리의 다른 글
| YOGTZE: Were the six letters left by Günther Stoll a clue (0) | 2026.07.27 |
|---|---|
| Hinterkaipek murder: Is it true that the killer stayed in a remote farmhouse (0) | 2026.07.27 |
| Merry Celeste: Why did the crew leave in a lifeboat (0) | 2026.07.27 |
| Oak Island Treasure: Between Money Pit Legend and Confirmed Evidence (0) | 2026.07.27 |
| Roanoke Colony: CROATOAN What One Word Says and Doesn't Say (0) | 2026.07.27 |