I discovered three extremely rare, unidentified, and unknown photos of Jūnisō from the 1860s. Read the story about this extraordinary find.
Shinjuku’s Lost Paradise, the 7-part essay about the history of Jūnisō Pond and Tokyo’s Nishi-Shinjuku, was completed last month. If you read along, you will have noticed that there were photographs of the shrine and the waterfall from the 1860s, but that there were no photos of Jūnisō Pond from before 1900.
Until now, the earliest known photos of Jūnisō Pond were from the 20th century, mainly postcards. Prints were mostly from the 1930s and later.
After an extensive search in photo collections worldwide I found two unidentified, unattributed, and undated photos of Jūnisō Pond in the collection of Nagasaki University Library. The one on top of this page, and the one below. It shows the same teahouses as in the top photo but from the other side of the pond.

As far as I have been able to ascertain neither photo has ever been published in a book, study, or online page about Jūnisō. This is the first time. Representatives of the Shinjuku Historical Museum and Jūnisō Kumano Shrine told me that they were unfamiliar with the photos. As did people who studied Jūnisō for many years. Additionally, a representative of Nagasaki University Library told me that no other researcher had contacted the library about these photos.
I also found an unknown photo of the shrine from the 1860s in the collection of Nagasaki University Library. More about that later.
Let’s first look at how I found, identified, and dated these photos.
The Discovery
Since I started my research about Jūnisō in early May I was repeatedly told that no photos of the pond existed from the 1800s. One person specifically told me that there were no prints from before the Showa Period, which began on December 25, 1926. There were postcards, but no prints.
I found five such prints, one from 1905/1906 and four from the second half of the 1910s—all of them extraordinary rare.1 All five are in the Duits Collection.
Perhaps there were earlier photos as well? Italian–British photographer Felice Beato (1833–1909) took photos of Jūnisō Kumano Shrine and Jūnisō Waterfall in the 1860s (see below). However, no Beato photos of the pond are known. To be absolutely sure, I meticulously combed Beato’s work. I found nothing.
By the 1860s, at least half of the many ukiyoe woodblock prints of Jūnisō showed the pond. Additionally, the pond was far more attractive to shoot than the waterfall. So why didn’t Beato shoot the pond?
Beato worked like a modern photojournalist and was a shrewd businessman. Being a photojournalist myself—with a career spanning three decades—it made no sense to me that Beato ignored the pond.
Perhaps he did take the shot, but it failed or had not survived. Or more intriguingly, it did survive and was both misidentified and misattributed. That would mean it was still out there, somewhere.
So, I redid my search at collections of vintage photos of Japan. Earlier searches had unearthed no 19th century photos of the pond when I used Jūnisō or its variations as a keyword. So this time I used keywords like pond, lake, teahouses, etc. In case of small collections I took no chances and checked every single photo.
I also inspected dealer collections. That is how I found the remarkable 1905/1906 photo introduced in PART 5.
After an extensive search I found the two photos introduced above. The screenshot below shows how the photographs were catalogued. The location, photographer, and date were unknown and there were no further explanations.

I continued searching in the library’s collection and next found a photo of Jūnisō Kumano Shrine from the same shelf. When comparing it with Beato’s photo of the shrine it became clear that it was the same place, but a different photographer.
I contacted the library and was told that except for the purchase date and the dealer’s name no additional information was known about the three Jūnisō photos.
The Nagasaki University Library is generally quite good at identifying the locations of the images in its collection. That the location of these photos remained unknown illustrates how extraordinarily difficult it was to identify them.
However, thanks to my research of Jūnisō I immediately recognized the pond. Especially a sketch by French military advisor Brunet and a painting by German painter Berg, introduced in PART 3 and shown below, were instrumental.
The three Jūnisō photos in the library’s collection are of similar size, seem to have been printed on the same paper, and show similar degradation. The two photos of the pond feature the same three people in the same clothes in the same teahouse, so they were almost certainly taken on the same day. All of this suggests that the three photos were likely taken by the same photographer.
A search for a lost photo by Beato had led me to three photos of the pond and the shrine. But by a different photographer. It turns out that two photographers had shot Jūnisō during the 1860s. An important discovery in itself.
Finding Proof
Although it was clear that these photos were of Jūnisō, it required conclusive evidence. First I made sure that it was not another pond. I checked if photos of other ponds resembled the pond in these two photos. None did.
Next I started comparing the individual items on the photos with the Brunet sketch and Berg painting. A friend assisted.
1. Tree
We first compared the unusual tree in the top photo with the one in Brunet’s sketch. The tree is shown from different angles, so it requires some three-dimensional thinking.

It is a very unusual tree, yet the form of the tree in the photo generally conforms with the one in the sketch. Especially interesting is that branches are cut off at apparently similar locations.
A large upward branch on the tree in the photo seems to be missing in the sketch. We didn’t think this was an issue. Perhaps Brunet took creative freedom. Or it was cut off, which would indicate that the photo predated the sketch.
2. Teahouses
We next compared the teahouses. The dimensions are the same. They are built the same. The roofs are the same. The paper lanterns are the same. The space between the teahouses on the right and the teahouse in the back agree.

The only real difference is the size of the teahouse in the back— it is narrower in the photo. This again could be explained by the photo predating the sketch and that the teahouse in the back was later extended.
3. Tree in the Back
One detail is especially persuasive. There is a tall tree in the back with the same form and at the same location. It is hard to notice on the sketch, but it is there.

I won’t bore you with all the comparisons and measurements that we took. Let’s have a quick look at the second image.
4. Second Pond Image
By comparing the teahouses on the two photos we were able to determine that this photo was taken from the other side of the pond. So you are looking at the area near the spot where Brunet sat to sketch his scene.

Two things are especially interesting. In the photo you can see a cluster of trees on the far right. They are tall and look quite different from the other trees. You can see a similar cluster of trees in the painting of the Prussian Expedition.
Also, there is a building in the back in both the photo and the painting, although it is a bit difficult to see in the photo.
There is one more interesting note to make. In both photos the same elderly man, boy, and girl can be seen in what appears to be the same clothing. So, these two photos were most likely shot on the same day.

Conclusive Proof
After all these careful comparisons I stumbled upon evidence that conclusively removed any possible doubt.
My research led me to a book about French military advisor Louis Kreitmann who visited Japan between 1876 and 1878. Titled Deux ans au Japon (1876-1878): Journal et correspondance de Louis Kreitmann, officier du génie, it features his diary, correspondence, and photos from five photo albums that he took back to France.2 It was published in 2015, relatively recent.
To my great surprise, the photo that resembled Brunet’s sketch is on page 38. The caption identifies the photo as the pond and teahouses at Jūnisō in Tokyo. Now there was 100% certainty that the two photos of Nagasaki University Library did indeed show Jūnisō Pond.
I contacted Japanese librarian Kaoru Baba who works with the publisher of the book, the Bibliothèque d’études japonaises in Paris. He introduced me to the Kreitmann family, who kindly allowed me to publish the page in their album:

Who is the Photographer?
As mentioned above, the three photos of Jūnisō held by the Nagasaki University Library have not yet been attributed. Louis Kreitmann took some of the photos in his albums himself, so Mr. Baba and I wondered if he was the photographer.
Based on his extensive research on Kreitmann, Mr. Baba was able to tell me that there was only one photo of Jūnisō in the Kreitmann Collection. I believe the three photos held by the Nagasaki University Library to be by the same photographer. So the absence of the other two photos in the Kreitmann Collection suggests that Kreitmann was not the photographer.
This hypothesis is supported by the approximate date of the photos. It is unclear whether the top photo does indeed predate the Brunet sketch (1867) as suggested above. However, there is another way to date the three Nagasaki University Library photos: comparing the photo of Kumano Jūnisō Gongen shrine with a photo of the shrine taken by Felice Beato. This has been dated to between 1863 and 1868 by the Getty Museum.
Although the photographer in the Nagasaki University Library photo stood much closer to the shrine and used a wider lens than Beato, there is a lot to compare. Especially noteworthy is a tree leaning over the pathway that looks remarkably similar in both photos.
This suggests that the two shrine photos were taken within several years of each other. That dates the three photos to the 1860s, or perhaps the very early 1870s. In other words, from before Kreitmann’s arrival in 1876.
Mr. Baba checked several other collections unavailable online and found no photos of Jūnisō. This rules out potential photographers and simultaneously confirms my conclusion that these three photos of Jūnisō are extremely rare.
For now, the three photographs remain unattributed.
Open Challenges
My research of the past seven months has extended the photographic history of Jūnisō by eight photos and about half a century.
What needs to be done next is to determine the photographer and more exact dates for the three photos from the 1860s.
Nagasaki University Library has no information about the Jūnisō photos beyond their purchase from Yushodo Bookstore (雄松堂書店) in 2013 (Heisei 25).3 Interestingly, the mount of one of the photos contains the Japanese text “Former Naidaijin (inner minister), Koga Michimasa” (久我前内大臣通誠).

Why is the name of Koga Michimasa (1660– 1719) on the mount? He died almost a century and a half before these photos were taken.
The Koga family (久我家, Koga-ke) is a branch of the powerful Minamoto clan that traces its descent from Emperor Murakami (村上天皇, 924–996). After the Meiji Restoration of 1868 the head of the family became a marquis (侯爵, kōshaku).
As it happens, family head Koga Takemichi (久我建通, 1815-1903) lived in the same era as the photos and was Naidaijin in 1862 (Bunkyū 2)—intriguingly close to the period that the photos were shot.
However, a representative of the Nagasaki University Library told me that the mount was likely reused and without relation to the Jūnisō photo.
I can see two ways to ascertain the photographer and more exact dates. One is to study photos with the same mounts as the Jūnisō photos (images 35 through 48 of shelf number 137 / catalog numbers 7454 through 7467).
The other is to discover additional prints of the Jūnisō photos. Perhaps they can be found in collections or books that I have not yet examined.
And … there may also be one or more unidentified photos of the pond by Beato still waiting to be discovered.
In short, this quest has not yet ended.
Thank You!
I want to thank all the readers that support Old Photos of Japan by subscription or donation. Though the bulk of my research is self-funded, your support helped make these discoveries possible. They belong to you as well. You have helped unearth a little piece of unknown Tokyo history. You truly make a difference!
This site is funded by readers like you
Old Photos of Japan provides thoroughly researched essays and rare images of daily life in old Japan free of charge and advertising. Most images have been acquired, scanned, and conserved to protect them for future generations.
I rely on readers like you to keep going. Please donate now to support this work.
Shinjuku's Lost Paradise
Find out more about the fascinating history of Jūnisō Pond and Nishi-Shinjuku, which is in effect the story of Tokyo itself.
Notes
1 To illustrate how rare these photos are, I have only found the photo with the rowboats in any publication or collection: in a Japanese collection and a book published in 1919: Morton-Cameron, W. H.; Feldwick, W. (1919) Present day impressions of Japan. London, The Globe Encyclopedia Company, 100.
An important note about these five photos. When I tried to locate the photos on maps in the Visual Chronology, the photos with the women did not fit, especially the photo with the two women. The building in that photo can be identified in later photos of Jūnisō Pond, but does not seem to appear on the other Enami photos or the 1905/1906 photo of the pond.
Possibly, the teahouses along the pond’s eastern shore were removed sometime in the late 1910s, and this building was built after their removal. Maps seem to confirm this hypothesis. This would suggest that Enami shot his four photos of Jūnisō Pond on two occasions, perhaps separated by several years.
2 Two Years in Japan (1876-1878): Journal and Correspondence of Louis Kreitmann, Engineering Officer. Currently, only a Japanese translation exists of this French book: 松崎碩子(2019)『クレットマン日記 : 若きフランス士官の見た明治初年の日本』東洋文庫.
3 Yushodo Bookstore merged with Maruzen in 2016 (Heisei 28); the name changed to Maruzen-Yushodo (丸善雄松堂).
Published December 18, 2024
Noel
Congratulations! That’s an amazing success for a “mere” seven month research. Reaching such an extensive analysis could often take years, not months. Love all the new information!
#000828 ·
Kjeld (Author)
@Noel: Thank you for your kind words, Noel, and good to hear from you again. You are right about studies like this often taking years. I wish I had more time to do more research!
#000829 ·