Quick answer: In Union of India v. Mohsin Sharieff, decided on 28 September 2026, a Division Bench of the Karnataka High Court set aside a single judge’s order that had restored a private purchaser’s name in the revenue records of land at Sonnappanahalli, Bengaluru. The Bench held that the land had been acquired in 1943 for the Yelahanka Air Force Station, that the Military Land Register is conclusive proof of the Defence Department’s title, and that a specific performance decree and court-executed sale deed obtained without the Defence authorities could not confer title on the buyer.
1. Introduction
Title disputes over land that the State acquired decades ago and then left undisturbed are a persistent feature of urban property litigation. The acquiring body’s records fade, survey numbers are re-cut, revenue entries drift, and a private claimant eventually appears with an agreement to sell from the original owner’s heirs. Union of India v. Mohsin Sharieff is a decision about exactly that sequence, and about which record wins when the paper trail of an acquisition is imperfect.
The Division Bench of D. K. Singh and H. Shanthi Bhushan JJ. considered land in Survey No. 33 of Sonnappanahalli village, where 3 acres 12 guntas had, according to the Union, been acquired in 1943 and used ever since as part of a taxi track for aircraft. The private claimant held a sale deed executed through a Court Commissioner in 2015 pursuant to a decree for specific performance. The single judge had set aside the revenue authorities’ orders removing his name from the records because the 1943 notification named Survey No. 30, not No. 33. The Division Bench reversed, holding that the discrepancy was explained by the rest of the acquisition record.
This article sets out the facts and procedural history, follows the Bench’s reasoning on identification of the land, the evidentiary force of the Military Land Register, and the effect of a civil decree on a stranger to it, and then assesses where that reasoning is strong and where it leaves questions open. It closes with what the decision means for due diligence on land with an acquisition history and for litigants who rely on court-executed conveyances.
2. Case summary and background
The claimant’s case. Mohsin Sharieff, the writ petitioner, said that one Thontarya owned Survey Nos. 32 and 33 at Sonnappanahalli, and that his legal heirs succeeded him. By an agreement of sale dated 8 October 1995, the heirs agreed to sell 4 acres 28 guntas in Survey No. 32 and 2 acres 11 guntas in Survey No. 33. He claimed to have paid Rs. 1,00,000 in advance and a further Rs. 33,95,000, but no registered deed followed. He sued for specific performance in O.S. No. 606/2008 before the Senior Civil Judge, Devanahalli, and obtained a decree on 26 August 2008. In Execution Petition No. 13/2014 a sale deed was executed through the Court Commissioner on 29 January 2015. His name was then entered in the revenue records, he paid property tax, and he said he had built a compound wall.
The Union’s case. The Union of India (Ministry of Defence), the Air Officer Commanding, Air Force Station Yelahanka, and the Defence Estates Officer said the land had been acquired under Sections 4(1) and 6(1) of the Land Acquisition Act, 1894, by notification in the Mysore Gazette on 12 February 1943, and that compensation had been paid to Thontarya. They accepted that Gazette Notification No. 3441 recorded the survey number as 30 but said this was an error for 33. They relied on a Surveyor’s letter of 31 May 1943, a survey report of 26 February 1944 recording that 3 acres 12 guntas of Survey No. 33 (total extent 4 acres 11 guntas) had been used for the aircraft taxi track, the Military Land Register (MLR), and mutation entries in Defence’s name.
The revenue proceedings. The Defence authorities challenged the petitioner’s mutation before the Assistant Commissioner under the Karnataka Land Revenue Act, 1964. By order of 18 July 2022 in R.A. (BNA) No. 425/2017-18, the Assistant Commissioner held the land to be Defence land and directed that the petitioner’s name be removed. The Deputy Commissioner dismissed the petitioner’s revision on 5 August 2023.
The single judge. In W.P. No. 12825/2024, the single judge allowed the petition on 21 July 2026, finding that Survey No. 33 did not appear in the 1943 notification, while Survey No. 32, described as the “Bajudar” of No. 33, did. The MLR entry could not fill that gap. The orders of the revenue authorities were quashed and the Tahsildar was directed to record the petitioner’s name, though the Union was given liberty to establish its title before the civil court. On 22 July 2026 the Union raised the petitioner’s two absolute sale deeds of 28 March 2019 in favour of one D. N. Lokesh. The single judge found no suppression because registration was completed only on 16 October 2024, after the writ petition was filed on 3 May 2024.
The holding. The Division Bench allowed the intra-court appeal and restored the revenue orders. Its central findings were these: the question was not whether “33” appeared in the notification but whether the whole acquisition record identified the acquired land with the petitioner’s Survey No. 33; the 1943 letter, the 1944 survey report, the compensation record, the village map and the MLR formed “one continuous chain”; the MLR is a public document carrying a presumption of correctness and, following Union of India v. Robert Zomawia Street, is conclusive proof of title; and the specific performance decree, being against parties who had no title and made in the absence of the Defence Department, could not create title. The Bench also drew adverse inferences from the petitioner’s conduct in later suing to cancel the 2019 sale deeds.
3. Legal analysis
3.1 Identification of the land: the record as a whole over a single defect
The most defensible part of the judgment is its treatment of the survey-number discrepancy. The single judge had, in effect, treated the Gazette notification as the exclusive source of what was acquired. The Division Bench held that the question was one of identification, to be answered from contemporaneous documents rather than later revenue entries. Because the acquisition was of 1943 and the agreement of sale of 1995, the Bench said, the earlier records “necessarily assume greater significance”.
That approach is orthodox. A notification is evidence of what the State intended to acquire; a wrong survey number does not necessarily change the land the boundaries, the survey report, the award and possession show was in fact taken. The Bench noted that the boundaries in the notification matched the land later identified as Survey No. 33, that the 1944 report records the extent used for the taxi track, and that the payment record shows Rs. 5,810 paid on 26 July 1945. Read together, these documents make the “typographical error” explanation credible, and it had been accepted by two levels of revenue authority.
One loose end deserves mention. The judgment records the Union’s pleaded case that the award was dated 28 February 1949, while the Additional Solicitor General’s submissions and the payment record refer to 26 July 1945. The Bench does not reconcile these. The discrepancy does not disturb the conclusion, since both dates lie decades before the 1995 agreement, but a party litigating an acquisition should expect the date of the award to be tested.
3.2 The Military Land Register: presumption or conclusive proof?
The Bench held that the MLR is a public document within Section 74(1) of the Bharatiya Sakshya Adhiniyam, 2023, as it forms part of the records of the acts of the Union, and that the Court may presume its entries correct under Section 119 as official acts regularly performed. It then relied on Union of India v. Robert Zomawia Street (2014) 6 SCC 707, where the Supreme Court, following Union of India v. Ibrahim Uddin, concurred that entries in the General Land Register maintained under the Cantonment Land Administration Rules are conclusive evidence of title. The Bench reasoned that the MLR is prepared by the Union to record land whose title vests in the Union and that the same principle applies, so that entries showing Survey No. 33 as acquired are “conclusive proof of the title of the Defence Department”.
Here the reasoning is thinner than the conclusion. Section 119 supports a rebuttable presumption; conclusive proof is a different evidentiary status, and the judgment moves between the two without saying which governs. The Robert Zomawia Street passage, as quoted, concerns the General Land Register under the Cantonment Land Administration Rules. The Bench treats the MLR as equivalent because both are Union-maintained registers of Union land, but it does not examine the statutory or regulatory basis on which the MLR is maintained or whether that basis confers comparable weight. The Bench also cites the Supreme Court’s decision in Union of India v. S. Narasimhulu Naidu, described as holding that MLR entries carry a presumption of correctness in the absence of evidence that the records were improperly kept. That formulation, a presumption displaceable by evidence, is closer to what Section 119 provides than to conclusiveness.
In practical terms the point may not have changed the result. The Bench also observed that the petitioner produced “no convincing material” that the land had ceased to be Defence land, so the presumption was unrebutted on any view. But the gap matters for the next case, in which a claimant may have material that does displace the entry. A court reading this judgment alone might conclude that no such attempt is permitted.
3.3 Nemo dat and the stranger to the decree
The doctrinal core is the Bench’s application of nemo dat quod non habet. If the land vested in the Union on acquisition and compensation was paid to Thontarya, his heirs had nothing to agree to sell in 1995, and no decree between them and the buyer could supply what they lacked. The Bench added that the Defence Department was not a party to O.S. No. 606/2008, and that a sale deed executed through the Court Commissioner could convey only such right as the judgment-debtors themselves possessed.
This reasoning is sound and consistent with basic principles. A decree for specific performance enforces a contract; it does not adjudicate title against the world. The Bench also invoked M. Venkatesh v. Commissioner, BDA (2015) 17 SCC 1, where the Supreme Court is quoted as holding that a sale after a preliminary notification is “void and non est in the eyes of the law”, leaving the vendee only a claim to compensation. That authority sits less comfortably with these facts, because the 1995 agreement came half a century after the notification and final acquisition, so the proposition adds little beyond nemo dat.
A subtler issue is the Bench’s characterisation of the 2008 decree as “collusive and an illegal decree” and a circumstance that “shocks the conscience of the Court”. The Bench relied on the gap of nearly thirteen years between the 1995 agreement and the 2008 suit, the fact that the decree was nevertheless granted, and a stated disparity between the consideration and the stamp duty on the 2015 deed. These are legitimate reasons for scrutiny, but the judgment does not record any finding of collusion in the civil proceedings, and the Bench did not treat the decree as void or set it aside. It held instead that the decree could not bind the Union. That narrower holding is the safer ground; the collusion language is better read as reinforcing it than as a separate ground. Nor does the judgment analyse limitation for specific performance, which runs from the date fixed for performance or from refusal, and not from the date of the agreement.
3.4 The writ court’s role and the closing of the title question
The Bench observed that the revenue authorities are competent to examine land records and identify land, and that the writ proceedings “could not have been converted into proceedings to determine the title”. Yet its conclusion, that the Defence title is conclusively established by the MLR, is a finding on title in all but name. The single judge had, as the judgment itself records, expressly left the Union free to establish title before the civil court. The operative part restores the revenue orders and does not mention that liberty.
That leaves an unresolved question: whether the petitioner remains free to seek a declaration in a civil court, and, if so, how a civil court would treat a Division Bench’s findings that the MLR is conclusive and that the decree was tainted. The petitioner had also filed O.S. No. 168/2024 for injunction, and the Union pointed to that suit as contradicting the writ petition’s statement that no suit was pending. The judgment leaves the fate of these suits to the parties.
3.5 Conduct and approbation-reprobation
The Bench spent considerable space on the petitioner’s later suit, O.S. No. 1094/2026, to cancel the 2019 sale deeds in favour of Mr. Lokesh. The petitioner there pleaded that the deeds were security for a loan of Rs. 5,64,40,000, while relying in the writ proceedings on title as owner. This “approbation and reprobation” was treated as relevant context. It is a fair point on credibility, though title to acquired land could not depend on it; the Bench’s real reasoning, in paragraphs 49 to 70, does not need it.
4. Practical significance
Due diligence on rural and peri-urban land. Buyers, banks and developers often verify title from revenue extracts (RTCs), mutation entries and tax receipts. This judgment is a reminder that these are fiscal and administrative records, not proof of title, and that an earlier acquisition can override them entirely. Where land lies near an airfield, cantonment, railway line or other public installation, diligence should include a request to the acquiring department for its land register extract, and a check of the original acquisition notification and award. A survey number appearing in a purchaser’s chain of title is not enough; the same land may appear under another number in an old Gazette.
Old agreements and delayed suits. A buyer relying on a decades-old agreement to sell, enforced years later by a decree, should expect that decree to be examined by any authority that was not a party. Pleadings should address how the vendors held title on the agreement date, and a plaintiff should consider joining the State or the department in whose favour an earlier acquisition may exist. A decree obtained without notice to that department is unlikely to protect the buyer.
Litigation strategy for the State. For departments, the judgment shows the value of preserving contemporaneous files: survey reports, award bills, payment records and registers. The Bench gave each document independent weight and treated them as one chain. Departments should also expect to be pressed on internal inconsistencies, such as the two dates for the award noted above.
Writ practice. The decision confirms that an intra-court appeal can reopen findings of fact in a revenue matter when the single judge is said to have relied too heavily on one document. It also sets a marker on disclosure: the Union pressed non-disclosure of the 2019 sale deeds, and although the single judge found no suppression because registration post-dated the petition, the Division Bench treated the later inconsistent stand as part of its overall assessment. Counsel should disclose transactions that are signed but not yet registered.
Revenue authorities and registrars. The judgment endorses the appellate revenue authorities’ power to examine acquisition records when deciding entries under the Karnataka Land Revenue Act. Registering authorities and Tahsildars processing mutations on court-executed deeds should take note that a mutation entry does not settle title and can be reversed on a showing of prior acquisition.
Cautions. This is a Division Bench decision of a single High Court, and its strongest statement, that the MLR is conclusive proof of title, rests on an analogy to the General Land Register that the judgment does not show the Supreme Court to have drawn in terms. Litigants who hold evidence that acquired land was released, denotified or never taken into possession should not treat the judgment as foreclosing that argument, because the Bench itself stressed that nothing credible had been produced by the petitioner to show the land ceased to be Defence land.
5. Conclusion
Union of India v. Mohsin Sharieff reaches a result that is easy to defend on its facts: land acquired for an operational Air Force facility, compensation paid, possession taken and recorded in the Union’s own register cannot be conveyed by the heirs of a man who lost title in the 1940s, and a decree in a suit to which the Union was a stranger cannot alter that. The Bench’s insistence on reading the acquisition file as a whole, rather than treating one wrong survey number as decisive, is a useful correction to a documentary literalism that can defeat legitimate public acquisitions.
The judgment is less secure where it elevates the Military Land Register from a presumption under the Bharatiya Sakshya Adhiniyam, 2023 to conclusive proof of title by analogy to a different register, and where it describes a civil decree as collusive without a finding to that effect. Those propositions may attract scrutiny in a later case with different facts, and their weight will depend on whether the Supreme Court or a larger bench addresses them.
For practitioners, the lessons are practical. Treat revenue entries as the beginning of title verification, not the end; look behind the survey number to the acquisition history; join the State when a decree may touch land it once acquired; and disclose every transaction touching the subject land when approaching a writ court. Whether the parties pursue the pending civil suits, and whether the judgment is carried further, are matters to watch.
Citations
- Union of India v. Mohsin Sharieff, Writ Appeal No. 1929 of 2026 (KLR-RES), High Court of Karnataka at Bengaluru (D. K. Singh and H. Shanthi Bhushan JJ.; judgment by H. Shanthi Bhushan J.), heard and reserved 9 September 2026, pronounced 28 September 2026. Source: Indian Kanoon, accessed 29 September 2026.
- Union of India v. Robert Zomawia Street, (2014) 6 SCC 707, Supreme Court of India, as cited and quoted in the judgment (paragraph 57), including its reference to Union of India v. Ibrahim Uddin, (2012) 8 SCC 148.
- M. Venkatesh v. Commissioner, Bangalore Development Authority, (2015) 17 SCC 1, as cited and quoted in the judgment (paragraph 64).
- Union of India v. S. Narasimhulu Naidu, Civil Appeal No. 2049/2013 c/w 13/2012, decided 27 August 2021, as cited in the judgment (paragraph 56).
- Air Officer Commanding, Air Force Station Yelahanka, Bangalore v. S. M. Ashwathanarayana, Civil Appeal No. 14815/2024, decided 17 December 2024, as cited in the judgment (paragraph 70).
- Prestige Lights Ltd. v. State Bank of India, (2007) 8 SCC 449, as cited by the Union on suppression of material facts (paragraphs 40 and 48).
Statutory provisions engaged
- Sections 4(1) and 6(1), Land Acquisition Act, 1894 (preliminary and final notifications of 12 February 1943).
- Sections 128, 136(2) and 136(3), Karnataka Land Revenue Act, 1964 (appeal and revision before the Assistant Commissioner and Deputy Commissioner).
- Sections 74(1) and 119, Bharatiya Sakshya Adhiniyam, 2023 (public documents; presumption of official acts regularly performed).
- Article 226, Constitution of India, and Section 4, Karnataka High Court Act (writ jurisdiction and intra-court appeal).
This article is not legal advice and does not create an attorney–client relationship. Readers should verify the certified copy of the judgment and any subsequent clarificatory orders before relying on it, and consult a qualified advocate on the facts of their matter.