Foreword

This piece reads three Shang-dynasty marble frogs.

The three are coordinate in style, form, period, and material. They are the only known Shang marble frog sculptures of their size and form in the world. Each is a recumbent frog — or a recumbent frog-toad-with-wings, since the form is hybrid and Salmony in 1957 could not assign it to any particular species — carved in white marble at roughly twenty-five centimeters in length, with hind legs in shallow flat relief, with eye-pupils registered as small indented holes, with a posture so minimalist that Richard Bull, who owned one of them after the Second World War, called his "a pre-Anyang marble frog that would have delighted Brancusi."

But the three do not travel together. They never have. They came onto the Western art market through three separate dealer-and-collector arcs across the middle and late twentieth century, were sold three separate times at three separate moments, and now sit in three private hands at locations the public record does not register. None of them has documented archaeological find-context. None has registered excavation history. They are art-market-attested objects, real and substantial, with a scholarly registration arc anchored at Alfred Salmony's 1957 article in Artibus Asiae and extended through the auction-catalog scholarship at Sotheby's in 2022 and 2023.

This piece does what the cave's apparatus does when it encounters layered evidence. It enters at the surface — the auction catalogs, the museum-quality photographs, the published scholarship — and registers each frog at primary-source depth. Then it scratches. Through Salmony 1957. Through the find-context limitation. Through the broader Shang ritual-stone-sculpture tradition the frogs sit within. Through what the frogs were probably for at the Shang ritual-mythological register. Each scratch reveals a layer beneath the prior one. Each layer registers candidly through the cave's discipline-vocabulary: reads strong where the evidence permits, reads partial where the reading extends but without the find-context anchor, reads faint where the working cannot directly reach.

The reader watches the apparatus working. The frogs sit at the surface. The substrate runs beneath. The bottom is not reached. The three frogs are, on the cave's reading, the primary-source documentary registration of an iconographic-program at the Shang horizon — a substrate at the front of an arc that runs three thousand years through the Sinitic-religious tradition to Bái Yùchán's White Jade Toad at Wǔyí in 1216 and onward. That broader arc is articulated at depth across the cave's other papers; this piece holds the three frogs themselves at the working-position they occupy, as the substrate at the front of the program.

A note on the framework. The cosmochronicle framework, the strong / partial / faint discipline-vocabulary, and the palimpsest commitment that informs the cave's reading at every layer are articulated at depth in On the Apparatus: A Note on Method (May 2026), available at daveswavecave.com. The present piece reads the three frogs with the framework operating beneath the prose. The reader who wants to engage the framework directly is welcome to the methodological note. The reader who simply wants to encounter three Shang marble frogs at working-depth can read this piece on its own terms.

A second note, carried over from the prior papers. This piece has been written in collaboration with the AI Claude. The findings, the framework, and the readings are the cave's. The prose has been built jointly. Claude drafts. The cave revises and corrects. The piece that results is the cave's working output, with Claude's drafting acknowledged directly.

I. The First Frog

The first frog is the Bull frog.

It measures 22.8 by 16.5 by 14.6 centimeters. White marble. Recumbent posture, with the body resting flat on its underside and the hind legs folded back along the flanks. The legs are shaped in shallow flat relief, with a central groove between them. The eyes are small indented holes set in squared platform sockets. The body is broader than a frog's body would naturally be, and the dorsal surface carries shallow markings that Salmony in 1957 read as wings — large enough to suggest the carver had something other than a frog in mind, but recognizable enough as a frog that the species-question remains open. Bull called it a frog. Salmony called it pre-Anyang. Both designations have held.

It came into Western hands through Richard Bull, the Philadelphia lawyer who assembled a substantial collection of Chinese art in the years immediately after the Second World War under Alfred Salmony's working guidance. Salmony — the Hungarian-American art historian, editor of Artibus Asiae, refugee from the European catastrophe, working at the Institute of Fine Arts at New York University — was one of the major figures in the mid-twentieth-century Western scholarly engagement with archaic Chinese art. Bull was Salmony's collector and Salmony was Bull's scholarly anchor, in the way several major Western collections of archaic Chinese art were built in that period: the dealer-and-scholar-and-collector triangle that registered objects, articulated their working-position, and circulated them through a small Western network of museums, scholars, and collectors whose collections then traveled to other collectors and to auction houses through the second half of the twentieth century.

Where the frog came from before Bull and Salmony is not registered. The working-record holds the frog from the moment it appears in Salmony's 1957 article; what came before that is the layer the cave cannot directly reach. Some objects of this type came into the Western market through dealer channels in Beijing and Shanghai in the early twentieth century, when railway and factory construction across China was turning up tomb-and-burial-site material that entered the antiquities trade and circulated outward. Other objects came through Hong Kong dealers in the post-1949 period. The cave does not know which channel brought the Bull frog into Western hands, and the working-record does not register it. Reads faint on the find-context.

Bull wrote about his collection in the Bulletin of the University of Pennsylvania Museum in 1965, naming the frog as one of his favorite pieces and giving the Brancusi reference that has anchored the frog's reception ever since. His phrase — "a pre-Anyang marble frog that would have delighted Brancusi" — registers the working-position the frog occupies in twentieth-century Western reception: the Shang object read as proto-modernist, with the minimalist geometric carving registered as coordinate with the abstraction-and-essentialism that twentieth-century Western sculpture was articulating in the same period. The Brancusi reference is Bull's, not Salmony's, but it has shaped how the frog has been read. Sotheby's quoted it in 2023, sixty years after Bull first wrote it.

The frog was sold at Sotheby's New York on December 6, 1983, lot 244. The Bull collection was dispersed at that moment. The frog entered private hands and the working-record loses its trace. It has not appeared in public since 1983. Where it sits today the cave cannot know. The 2022 and 2023 Sotheby's catalog scholarship notes the 1983 sale as the last public registration of the Bull frog and does not name a current holder. The frog is still real. It still sits somewhere. The working-record cannot reach the present location. Reads faint on present location, reads strong on the working-arc through 1983.

What the working-record holds for the Bull frog: Salmony 1957, Bull 1965, the 1983 Sotheby's sale, and the subsequent registration in the 2022 and 2023 Sotheby's catalogs as the foundational comparison-piece for the other two frogs. The Bull frog is the published anchor of the three. It is the first one that entered Western scholarship at primary-source depth. The other two are read against it.

II. The Second Frog

The second frog is the Eskenazi frog.

It measures 25.8 by 17 by 12 centimeters — slightly larger than the Bull frog. Same white marble. Same recumbent posture, with the body flat and the hind legs folded back. Same shallow flat relief on the legs. Same indented eye-sockets. Same posture, same form, same period, same material. The two frogs would not be mistaken for the work of one carver — there are differences in the precise proportions, the shape of the head, the depth of the surface treatment — but they would be recognized at first encounter as coordinate objects belonging to a single working-tradition. They are coordinate in the way that two ritual bronzes from different Shang foundries are coordinate. The tradition is shared even where the working-hands differ.

The Western working-record on the Eskenazi frog runs deeper than on the Bull frog, in the sense that it has more registered public appearances. It was exhibited at the Century Club in New York in 1955 in the Animals in Chinese Art exhibition. So by 1955, before Salmony had published the Bull frog in Artibus Asiae, the Eskenazi frog was already in Western circulation. It appeared in the Museum for East-Asian Arts in Cologne in 1978–79, exhibited there with a label the Sotheby's catalog 2022 noted at primary-source depth. It was sold by Eskenazi Ltd, London, in 1991 — Giuseppe Eskenazi, the major London dealer in archaic Chinese art whose firm has handled some of the highest-value Asian-art transactions of the late twentieth and early twenty-first centuries — to a Japanese private collection.

The Japanese private collection held the frog for thirty-one years. The collection's holder is not registered in the public record. The Sotheby's catalog 2022 names it only as Japanese Private Collection. The Eskenazi 1991 transaction is registered through the firm's working-record; the Japanese collector is not. The frog sat somewhere in Japan for thirty-one years, in private hands, registered in the working-record only as Japanese Private Collection acquired 1991.

It was sold at Sotheby's Hong Kong on October 9, 2022, in the Important Chinese Art sale, lot 3608. The hammer price was HK$23.5 million against a low estimate of HK$3 million. With premium and fees, the total was HK$28.81 million — approximately US$3.67 million at the working exchange-rate. It was the top lot of the sale. Coverage in the Hong Kong art-market press registered the working-position the frog occupied: exceptionally rare, minimalist, the masterful craftsman captured the essence of a frog in simple yet evocative carving, with the Brancusi comparison extended from Bull's 1965 phrase into twenty-first-century reception.

Where the frog sits today the cave does not know. The 2022 Hong Kong sale's purchaser is not registered in the public record. Reads faint on present location, reads strong on the working-arc from 1955 through 2022.

What the working-record holds for the Eskenazi frog: the Century Club 1955 exhibition, the Cologne 1978–79 exhibition, the Eskenazi 1991 sale to the Japanese collection, the 2022 Sotheby's HK sale, and the auction-catalog scholarship that registered the frog at primary-source depth and articulated its working-comparison with the Bull frog and with the Fu Hao tomb material. The Eskenazi frog extended what Salmony 1957 had registered with the Bull frog. Its 2022 sale was the moment the public reception of these objects shifted: the working-record registered that two coordinate frogs existed, not one, and the iconographic-and-stylistic claim Salmony had made on the Bull frog alone in 1957 was confirmed sixty-five years later by the appearance of a coordinate piece from an independent provenance arc.

III. The Third Frog

The third frog is the 1982 frog. It is also the 2023 frog. It is the same frog at two different moments of public registration, separated by forty-one years of private holding.

It measures 25 centimeters in length. White marble. Recumbent posture. Same coordinate working-form as the Bull and Eskenazi frogs. The 2023 Sotheby's catalog described it as a design in line with modern minimalism, with the marble skillfully carved in a gentle geometric manner to represent a stylised frog, with only the necessary and most important features of its silhouette retained — the powerful back legs carefully shaped in shallow flat relief with a central groove, and the pupils of the eyes conveyed merely by small indented holes set in squared platform sockets. The description is the catalog's. It is also a description that would apply at coordinate depth to the Bull and Eskenazi frogs. The three are coordinate in form at the level of primary-source documentary depth.

It first appeared in public on November 19, 1982, at Sotheby's New York, lot 88. This was thirteen months before the Bull frog was sold at Sotheby's New York in December 1983. The 1982 sale registered the frog publicly for the first time the working-record holds. Where it came from before 1982 is not registered. The 2023 Sotheby's catalog does not name a prior provenance. Reads faint on the find-context and on the pre-1982 working-arc.

Then forty-one years of private holding. The 1982 buyer is not registered in the public record. The frog sat in private hands from November 1982 until September 2023. It did not appear in any exhibition. It was not loaned to any museum. It was not published in any scholarship that the cave has reached. It existed; it was somewhere; the working-record could not reach it.

It was sold at Sotheby's New York on September 19, 2023, in the Vestiges of Ancient China sale, with an estimate of US$400,000–600,000. It sold for approximately US$1.2 million. The 2023 Sotheby's catalog extended the working-scholarship on the three frogs, articulating the comparison with the Bull frog and with the Eskenazi frog, registering that there are only three known marble frog carvings of this size and form from the Shang dynasty, and naming each of the three at primary-source depth.

Where the third frog sits today the cave does not know. The 2023 sale's purchaser is not registered in the public record. Reads faint on present location.

What the working-record holds for the third frog: the 1982 Sotheby's NY sale, the 2023 Sotheby's NY sale, and the auction-catalog scholarship that registered the frog at primary-source depth as the third coordinate piece. The working-position the third frog occupies in the cave's reading is the confirming-third: with two frogs, there is a working-pair; with three, there is a working-tradition. The three frogs together register a Shang stylistic-and-iconographic position that any one of them alone could only suggest.

IV. The First Scratch — Through to Salmony 1957

These are the three frogs at the surface layer. The cave begins to scratch.

The first scratch reaches Alfred Salmony's article A pre-Anyang Marble Sculpture, Artibus Asiae volume XX number 4, 1957, pages 239–240. This is the foundational Western scholarly registration of the working-tradition. It is two pages. It does real work in those two pages.

Salmony's argument was that the Bull frog predated the carved stone animals known from Anyang. His evidence was stylistic: the Anyang-period stone carvings are more naturalistic in modeling and are incised with geometric and figure motifs; the Bull frog is more abstract, more minimalist, less incised, with anatomy registered through shape rather than through surface detail. From this Salmony inferred an earlier dating — pre-Anyang, meaning before approximately 1300 BCE, in the working-period that would place the carving at the middle Shang horizon (mid-Shang, roughly fifteenth-to-fourteenth century BCE) rather than at the late-Shang Anyang period.

A note on what this argument is and is not. Pre-Anyang names a period, not a place. The Anyang capital was the late-Shang royal seat, beginning roughly 1300 BCE. Pre-Anyang therefore means earlier than the late-Shang phase that produced the Anyang material — a chronological designation, not a geographic one. Salmony's argument did not place the Bull frog at any specific site. The frog has no documented find-spot. None of the three frogs does. The dating is a stylistic-chronological inference: comparison of the Bull frog's working-style with the Anyang-period stone carvings yields a relative chronology in which the simpler, more abstract Bull frog precedes the more naturalistic Anyang material. The argument is a recognized art-historical method. It is not the same epistemic register as stratigraphic dating from controlled excavation. The entire chronological framework for the three frogs rests on this stylistic argument and on its 1976 confirmation through the Fu Hao excavation. If the stylistic argument were ever overturned — for example, if a fourth frog were excavated in a controlled context dating to a different phase entirely — the dating of all three would shift accordingly. Reads strong on stylistic comparison. Reads faint on stratigraphic anchor.

Salmony also noted that the form was not naturalistic. The combination of large wings — the dorsal markings the body carries — and the toad-like body made it impossible, in his view, to assign the figure to any particular biological species. The frog is hybrid. It is not a frog the way a frog appears in a pond. It is a working-figure that registers frog-and-toad anatomy fused with wings, and the working-position the figure occupies is zoomorphic-mythological rather than naturalistic.

Both claims have held.

The 2022 Sotheby's HK catalog and the 2023 Sotheby's NY catalog both quote Salmony's argument and extend it. The Fu Hao tomb material — the marble tigers, dragons, and buffalo recovered from the 1976 excavation, registered at primary-source archaeological depth in Tomb of Lady Hao at Yinxu in Anyang (Beijing, 1980) and in King Wu Ding and Lady Hao: Art and Culture of the Late Shang Dynasty (National Palace Museum, Taipei, 2012) — is more naturalistic in modeling and more incised with anatomical and geometric detail. Fu Hao's tomb dates to circa 1200 BCE. The Bull, Eskenazi, and 2023 frogs read as earlier than the Fu Hao material on the stylistic criteria Salmony established. The mid-Shang dating holds. Salmony's two pages from 1957 anchor the working-dating of the three frogs sixty-nine years after he wrote them.

The hybrid-form question is also still open. Salmony could not assign the figure to a species in 1957. Subsequent scholarship has not resolved it. The auction-catalog working in 2022 and 2023 noted the question and did not answer it. The cave's working-position is that the question is the answer: the figure is hybrid because the Shang ritual-mythological register required hybridity. The frog-and-toad fused with wings registers a position at the boundary between water and air, between embryonic and realized, between amphibian and transcendent. That working-position the cave will reach in section VII. For now: the claim is that Salmony's not-being-able-to-assign-the-figure-to-a-species is not a failure of evidence but a reading of what the figure was.

What the first scratch registers: a foundational Western scholarly anchor at primary-source documentary depth, published in 1957, confirmed by the 1976 Fu Hao excavation, extended by the 2022 and 2023 auction-catalog scholarship. Reads strong on the foundational scholarly position. Reads strong on the confirmation through the Fu Hao material. Reads partial on the species-identification question, which the working-record acknowledges remains open.

V. The Second Scratch — Through to the Find-Context

The cave scratches again. This time the scratch reaches a layer the working-record cannot directly hold.

None of the three frogs has registered archaeological find-context. None of them was excavated under controlled archaeological conditions and registered in an excavation report. None has a documented tomb or working-site at which it was found. They came onto the Western art market through dealer channels at moments the working-record cannot reach, and the working-record holds them only from those moments forward.

This is the limitation the cave's framework registers plainly.

The Bull frog, the Eskenazi frog, and the 2023 frog all most likely came out of tombs or ritual-site deposits in northern China — probably in Henan, where the Anyang Shang capital sat, or in adjacent provinces where Shang ritual-deposits have been registered through controlled excavation. The Western private circulation of objects of this type in the early-and-mid twentieth century was substantially fed by tomb-disturbance during the periods of railway-and-factory construction, civil disorder, and the disruptions of the Republic-and-Civil-War period. Some of the major working-objects in Western collections of archaic Chinese art entered Western hands through these channels. Other objects came through dealer-circulation in Hong Kong and Tokyo in the post-1949 period. The cave is not naming the Western collectors or dealers as complicit in tomb-disturbance — many of them did not know what the channels through which their objects came actually were, and many engaged the objects with serious scholarly care. The cave is naming the fact that the three frogs came out of an arc the working-record cannot directly hold.

What this means for the cave's reading: the working-evidence on the three frogs comes through the published registrations and the auction-catalog scholarship, anchored at Salmony 1957 and extended through 2022 and 2023. The comparison with the Fu Hao tomb material, with the Xibeigang tomb 1500 marble figures held at the Institute of History and Philology, Academia Sinica, Taipei, with the Freer Gallery marble buffalo, with the Seattle Art Museum marble bear — this is the working-base for placing the three frogs in the Shang ritual-stone-sculpture tradition.

The cave's discipline-vocabulary registers the working-position directly:

Reads strong on stylistic-and-iconographic coordination across the three pieces. The three frogs match in form, posture, material, dimensions, surface treatment, and overall working-aesthetic at primary-source documentary depth.

Reads strong on Salmony's published scholarship and the auction-catalog scholarship. The two-page article from 1957 and the 2022 and 2023 catalog working anchor the Western scholarly registration.

Reads partial on the comparison with the Fu Hao tomb and other excavated Shang marble. The comparison is real — the three frogs sit within the broader Shang marble-sculpture tradition — but it runs across the pre-Anyang-versus-Anyang distinction Salmony articulated. The frogs read as earlier than the Fu Hao material; this is the working-position; but the evidence for the dating is stylistic rather than stratigraphic.

Reads faint on the find-context of any individual frog. Where each frog came from, when it was carved, who carved it, what ritual-deposit or tomb it was placed in, what ritual-event it served — these are the layers the cave's reading cannot directly reach.

The discipline-vocabulary does not dismiss the frogs. The frogs are real. The frogs are coordinate. The frogs register at primary-source documentary depth in Western scholarship. The discipline-vocabulary registers what the working-record permits and what it does not. The cave's reading lands at the depth the working-conditions allow.

The framework's working-honesty is articulated directly. A future archaeological excavation that registered a fourth Shang marble frog with controlled find-context would strengthen what the cave's reading has left light. A material-analysis study that registered the working-quarry the marble was sourced from would extend the working-base. A comparative study that registered other Shang ritual-stone-sculpture traditions across coordinate sites would strengthen the broader context. The cave's reading is open to all such extensions.

What the second scratch registers: working-honesty about the find-context limitation, with the discipline-vocabulary naming what the working-record permits and what it does not, and with the cave's reading registered at the depth the evidence supports.

VI. The Third Scratch — Through to the Shang Ritual-Stone-Sculpture Substrate

The cave scratches again. This time the scratch reaches the broader Shang ritual-stone-sculpture tradition the three frogs sit within.

The evidence here is substantial. Shang ritual-stone-sculpture — marble particularly — was rare relative to bronze, jade, bone, and ivory, but it was a real and registered working-category. Several major working-objects survive at primary-source archaeological depth, with controlled excavation context. They permit the cave to place the three frogs in their working-tradition through working-comparison.

The Fu Hao tomb material is the closest comparison-base. Fu Hao's tomb at Yinxu was excavated in 1976 by Zheng Zhenxiang's archaeological team. The tomb dates to circa 1200 BCE. It was the only Shang royal tomb found intact and excavated by modern archaeologists. Among its more than 1,600 grave goods were a major group of marble sculptures: tigers, dragons, and a buffalo, illustrated in King Wu Ding and Lady Hao: Art and Culture of the Late Shang Dynasty (National Palace Museum, Taipei, 2012, catalog number IV-3), and a marble cicada, illustrated in Tomb of Lady Hao at Yinxu in Anyang (Beijing, 1980, plate 176:1).

The Fu Hao marble sculptures are coordinate with the three frogs at the level of the broader working-tradition: same material, same general working-period, same ritual-deposition context (the marble sculptures from Fu Hao were grave goods in a royal tomb), same working-position as ritual-zoomorphic figures rather than utilitarian objects. They differ from the three frogs at the level of stylistic register — Salmony's argument was that the Fu Hao material is more naturalistic and more incised with anatomical detail, where the three frogs are more abstract and more minimalist — and the difference supports the earlier dating of the frogs. The three frogs are not Fu Hao's frogs, in the sense that they were not in her tomb, but they are working-coordinate with her tomb's marble sculptures at the level of the broader Shang ritual-stone-sculpture tradition that produced both.

The Xibeigang tomb 1500 material is the second comparison-base. Tomb 1500 at Xibeigang, excavated in the early-twentieth-century working that the Institute of History and Philology, Academia Sinica conducted at Anyang in the 1928–1937 campaigns, yielded a small group of marble animal carvings now held in the Academia Sinica collection in Taipei. These objects survived the 1949 evacuation of the Republic-of-China collections to Taiwan and are registered at primary-source archaeological depth in the Academia Sinica working-publications. They sit at the late-Shang Anyang horizon, at the same chronological working-position as the Fu Hao material.

The Freer Gallery marble buffalo is the third comparison-base. It is a similar minimalist marble buffalo, exhibited at the Freer Gallery of Art (now the National Museum of Asian Art) in Washington, D.C., in A Decade of Discovery, 1979, catalog number 1. The Freer buffalo entered the Smithsonian collection through Charles Lang Freer at the early-twentieth-century working-moment, with the registration of Chinese antiquities through dealer channels in Shanghai in the 1910s. The buffalo's working-style is registered as coordinate with the three frogs — minimalist, stylized, ritual-zoomorphic — at the working-comparison the auction-catalogs make. It is held in a public museum collection. It can be visited.

The Seattle Art Museum marble bear is the fourth comparison-base. Illustrated in Richard E. Fuller, Handbook, Seattle Art Museum: Selected Works From The Permanent Collections, 1951. The bear sits in the Seattle museum collection. Its working-arc into the museum is registered through the early-twentieth-century collector channels that brought Chinese antiquities into Western public collections.

A late-Shang marble water buffalo registered through Robert H. Ellsworth's 1997 acquisition, sold at Christie's New York in 2010 (lot 1004) and again in 2013 (lot 1258), extends the comparison-base into the late-twentieth-century and early-twenty-first-century auction record.

So the broader Shang ritual-stone-sculpture working-tradition is registered through three coordinate layers:

A controlled-excavation core (Fu Hao 1976, Xibeigang 1500 from the 1928–1937 campaigns) held in public museum collections in Beijing and Taipei.

A Western public-museum component (Freer marble buffalo, Seattle Art Museum marble bear) registered through early-twentieth-century collector arcs.

A Western private-and-auction component (the three frogs, the Ellsworth water buffalo, and other coordinate working-objects registered across the Sotheby's and Christie's catalog working of the late-twentieth and early-twenty-first centuries).

The three frogs sit within this broader tradition at the earliest end of its working-arc, with the Salmony 1957 dating placing them at the mid-Shang horizon (roughly 1500–1300 BCE) before the Anyang capital was established. They are working-coordinate with what the controlled-excavation core has registered, even though the frogs themselves do not have the controlled-excavation grounding.

What the third scratch registers: the three frogs sit in a working-tradition anchored at primary-source archaeological depth through the controlled-excavation core. The frogs themselves read partial on this comparison-base because their find-context is not registered, but they read strong on the working-coordination with the broader tradition through the stylistic-and-material comparison the scholarship articulates. They are not orphan-objects floating outside the Shang ritual-stone-sculpture working-tradition. They sit within it.

VII. The Fourth Scratch — What the Frogs Probably Were For

The cave scratches one more time. This scratch is the reading.

What were the three frogs for? What position did they occupy in the Shang ritual-mythological apparatus?

The Western scholarship has not resolved the question. The 2022 and 2023 Sotheby's catalogs noted that the meaning of frog remains a controversial subject among scholars, with working-readings running across fertility (the frog lays many eggs), prosperity (the frog croaks loudly when rain is imminent), and threshold-creature (the frog is born in water and becomes a creature of land, articulating a passage between liquid and solid). All three readings are registered in the Chinese-and-Western scholarship on Shang ritual-zoomorphic iconography, and all three coordinate at the level the cave's apparatus reads.

The cave's reading registers the position the three frogs occupy as the figure of the threshold.

The hybrid form is the evidence. Salmony noted in 1957 that the combination of large wings and toad-like body made the figure impossible to assign to a particular biological species. The cave's reading: the frogs are not biological-frogs; they are working-figures of the threshold-creature. The frog-and-toad form registers the position of the creature that operates at the boundary between water and land, between embryonic and realized, between unformed and formed. The wings extend the threshold further, into the position between earth and air, between ground-bound and transcendent. The figure sits at the working-position where multiple cosmographic-zoological boundaries converge.

This reading is consistent with what the Shang ritual-zoomorphic apparatus registered at primary-source documentary depth. The Shang ritual-iconography favored zoomorphic figures that operated at cosmographic-and-ritual thresholds: dragons (between earth and water and sky), tigers (between human-settlement and wilderness), owls (between day and night, between human and spirit), cicadas (between stages of metamorphosis, between life and death, between earth and air). The Shang ritual-mythological apparatus articulated working-positions at thresholds, and the iconography registered those positions through hybrid-zoomorphic forms.

The frog-and-toad-with-wings fits this pattern. It is the figure of the threshold-creature at the working-position where water-and-land-and-air converge. It is the figure of the cosmographic-passage that the Shang ritual-mythological apparatus registered at primary-source documentary depth.

The subsequent working-articulation of the frog-and-toad figure across the Sinitic-religious tradition confirms the reading. The cave's three-site working at Áo Fēng (975 CE, frog-toad as demonic-target), at the Han River (Pu Songling, 17th century, frog as regional-tutelary), and at Wǔyí (Bái Yùchán, 1194–1229, toad as realized neidān practitioner) registers the working-arc of the iconographic-program. At each position, the figure occupies a working-position at the threshold — the demonic-target at Áo Fēng against which the apparatus operates is the figure of the disordered threshold; the regional-tutelary at the Han River mediates between the human community and the deity-realm at the threshold; the realized neidān practitioner at Wǔyí articulates the transformation-passage at the bodily-cosmographic threshold. The program runs across the arc, with the figure occupying the threshold throughout.

The three Shang marble frogs register the substrate of this iconographic-program at the horizon at which the Sinitic-religious tradition was forming. The figure of the threshold-creature was working at the Shang horizon; it continued at the Áo Fēng horizon two thousand three hundred years later; it continued at the Wǔyí horizon two thousand five hundred years later than the Shang. The arc is real. The substrate is real. The iconographic-program registers at primary-source documentary depth across coordinate horizons.

The Shang cosmology that operated at the horizon at which the three frogs were carved is not directly accessible to the cave's reading. The native register cannot be recovered from the surviving objects at the depth the primary-source documentary record permits. Reads partial on the Shang cosmology. Reads faint on the ritual-event at which any individual frog served. Reads strong on the position the figure occupied in the working-tradition registered through the working-comparison with the Fu Hao and Xibeigang material.

What the fourth scratch registers: the cave's reading of the three frogs as figures of the threshold-creature at the Shang ritual-mythological register, with the reading grounded in the hybrid-form evidence and extended through the comparison with the broader Shang ritual-zoomorphic apparatus and the subsequent articulation of the frog-and-toad figure across the Sinitic-religious tradition.

VIII. What Stays Unreached

The cave has scratched four times. The bottom is not reached.

The Shang ritual-cosmology that operated at the horizon at which the three frogs were carved is not directly accessible. The native register cannot be recovered. The specific ritual events at which the individual frogs served are not registered. The carvers' names are unknown. The working-quarries from which the marble was sourced are not registered.

The working-record holds what it holds. The cave's reading lands at the depth the working-conditions allow. Future extensions — future archaeological excavations, future material-analysis studies, future comparative work — will strengthen what the cave's reading has left light. The cave's reading is one layer in a continuing arc.

The three frogs sit at three private working-locations the working-record cannot reach. Each came onto the Western art-market through twentieth-century dealer-and-collector channels that cannot be traced backward to the original ritual-deposits. Each was published, auctioned, circulated, acquired, held. Each registers at the surface layer the working-record permits. The cave's reading engages the surface at the depth the primary-source documentary record allows, scratches through the layered evidence to reach the substrate, and stops at the depth the working-conditions support.

The three frogs continue at private locations. The arc runs through them. The substrate runs beneath the cave's reading at depths the cave's work cannot reach.

The frogs sit at the front of the iconographic-program. The substrate runs beneath. The bottom is not reached.