Grotius’s Position on Implied Servitudes by Means of Destinatione Patris Familias


According to Grotius in his Inleiding (2.36.6), the actual use of two houses by the same owner could lead to the implied grant of a servitude if he transferred one of the houses to someone else, ‘without any mention either the one way or the other’. Various interpretations of this text exist, but the consensus is lacking. In this article, the author investigates the meaning and influence of Grotius’s position on implied servitudes in both his time and the following centuries. This research shows how Grotius’s opinion progressed from Bartolus’s approach to implied servitudes and sheds new light on the creation of servitudes by means of destinatione patris familias in the Netherlands.


Introduction
In his Inleiding tot de Hollandsche Rechts-geleerdheid (Introduction to Jurisprudence of Holland) (1631), Hugo de Groot discussed several different ways of creating servitudes. He remarked that a servitude could be created by agreement (toezegging), followed by tolerance ('t gheheng), by will, and by prescription.1 He then suggested that the creation of servitudes did not necessarily require an express agreement but that the following facts sufficed: Zoo wanneer een vollen eigenaer van twee huizen het eene huis heeft ghebruict met eenige gherechtigheid over het andere huis: ende daer nae door opdracht de huizen in eigendom werden gescheiden, zonder iet uit ofte in te zegghen, zoo behoud yeder huis zijne voordeelighe ende nadeelige gerechtigheden zulcs als te vooren. So, when the full owner of two houses has used the one house with some entitlement to the other house, and afterwards the houses are divided in ownership by transfer, without any mention either the one way or the other, then each house will keep its advantageous and disadvantageous entitlements as before.2 According to Grotius, the actual use of one house in a servient way and the other house in a dominant way by the same owner had legal effects. It led to the creation of a servitude if the owner of both houses transferred one of the properties to someone else.
Van Leeuwen and Van der Keessel have argued that Grotius's opinion was limited to cases where a servitude was previously lost by a merger (confusio).3 A merger occurred when the owner of a dominant estate became the owner of the servient estate (and vice versa). According to Van Leeuwen and Van der Keessel, Grotius had meant that the former servitude revived if the owner of both houses transferred one of the houses to someone else. Groenewegen van der Made seems to have had the same interpretation of Grotius.4 Voet, however, suggested that Grotius's position was comparable to the creation of servitudes by means of destinatione patris familias.5 This concept was developed by Bartolus and applied to cases where the owner of two estates used one of his estates in a servient way in favour of his other estate and transferred the actually servient estate. According to Bartolus, this led to the implied grant of a servitude if its cause was continuous and permanent. 6 Voet, however, did not elaborate on the extent of the similarity between Bartolus's approach and that of Grotius's. Voet rejected Grotius's opinion by arguing that a seller could only create a servitude by reserving it expressly in the contract.7 At the end of the eighteenth century, Didericus Lulius, Pieter and Reinier van Spaan, and Joannes van der Linden discussed Grotius's position. Still, they did not restrict it to the revival of lost servitudes or the creation of a servitude if its cause was continuous and permanent.8 In the South African case, Salmon v Lamb's Executor & Naidoo (1906), judge Kotzé, however, remarked that it was the 'generally-received interpretation' that Grotius's position only applied to the revival of servitudes after a merger.9 Not only is consensus on the interpretation of Grotius lacking, but it is also unclear what Grotius's influence was on the recognition of implied servitudes in his time and the following centuries.10 Eighteenth-century Dutch legal scholars remarked that Grotius's opinion was corroborated by the local laws of the Middelburg, Vlissingen, Schoonhoven, and Dordrecht.11 Moreover, the creation of servitudes by means of destinatione patris familias was eventually adopted in the French civil code of 1804 and the Dutch civil code of 1838. The development of implied servitudes in Italy, France, and other European countries12 has received ample scholarly attention, but this development in the Netherlands (and Grotius's part in it) until the nineteenth-century civil codes remains a mystery.
In this article, I will investigate the meaning and influence of Grotius's position on implied servitudes in both his time and the following centuries. I will first discuss the Roman roots of Grotius's remarks on the creation and merger of servitudes. Consecutively, I will analyse how Roman jurists and the medieval scholar Bartolus dealt with the implied grant of servitudes. I will then discuss how Bartolus has influenced early modern local laws and, after analysing the various interpretations of Grotius's position on implied servitudes, argue that Grotius progressed from Bartolus's approach to implied servitudes. Finally, Grotius's influence on the recognition of implied servitudes in the Netherlands will be investigated, both in his time and the following centuries.

The (re)creation and Merger of Servitudes in Roman Law
Grotius's standard ways of creating servitudes -agreement (toezegging), followed by tolerance ('t gheheng), will, and prescription -clearly echoed Roman law.13 In the Institutes, emperor Justinian expressly mentioned that servitudes were created by will or informal agreements and stipulations.14 Moreover, the prescription of servitudes by means of praescriptio longi temporis was recognised.15 The jurist Ulpian also mentioned the (praetorian) protection of someone who was simply granted access to land by its owner and who was tolerated to use the land in a servient way,16 and of someone who enjoyed the long use and quasi-possessio of a servitude.17 The glossators and commentators were reluctant to accept that servitudes could be created by mere informal agreements and stipulations. They argued that a quasi traditio was required, 'such as by means of tolerance' (ut per patientiam).18 This requirement was in accordance with their position on the requirements for a transfer of ownership. They considered C. 2.3.20 to be the general rule, which required a traditio for the transfer of ownership.19 Despite the fact that Grotius had argued from a natural law perspective that a transfer of ownership required no delivery,20 he did consider tolerance ('t gheheng) to be required for the creation of servitudes.
Grotius was unwilling to recognise the requirement of a quasi-delivery before a local court. According to Grotius in one of his legal opinions, the imperial decree that required that transfers of immovables be passed before a local court did not apply to the creation of servitudes, since servitudes did not qualify as transfers (servitutis imposition non comprehenditur sub alienatione).21 Dig. 7.1.6.1 (Gaius 7 ad edictum provinciale) concerned a right of usufruct but Gaius stated that servitudes were created in the same way as rights of usufruct. See Voet agreed with Grotius that servitudes were created by agreement and quasi-delivery, 'which consists in use and sufferance or in the pointing out of a spot for the exercise of the servitude' , but rejected the notion that local customs required no quasi-delivery before the local court.22 Groenewegen van der Made, Van Leeuwen, and Van der Keessel were of the same opinion.23 In the Digest, a frequently mentioned way to create a servitude was the reservation of a servitude when an estate was transferred. The transferor could encumber the transferred 'servient' estate by reserving a servitude in favour of his own 'dominant' estate at the moment of the transfer. Julian stated: A man who has two tracts of vacant ground can, on the conveyance of one of them, impose a servitude on it in favour of the other.24 Some texts suggest that this reservation was made in the contract for the transfer, whereas other texts suggest that it was made at the moment of the transfer.25 Marcian argued that a transferor could compel the transferee to create a servitude if the parties had agreed upon the creation of a servitude in the sale and the transferor had forgotten to reserve the servitude at the moment of transfer.26 In classical Roman law, a reservation of a servitude was implemented in the required formal declaration for a mancipatio or in iure cessio.27 In Justinian law, it would not have mattered if the parties agreed on the reservation of a servitude in the contract or at the moment the property was delivered. Justinian law required a iusta causa and a delivery (traditio) in order for a transfer of ownership to take place. If the parties had agreed in the contract 22  that the transferor could reserve a servitude, then the transferee only received ownership encumbered by a servitude at the moment the property was delivered.28 If the parties agreed on the reservation of the servitude at the moment of delivery, the agreement could be seen as an amendment to the iusta causa.
A transferor could not only create a servitude when transferring a 'servient' estate, but he could also encumber his own estate in favour of the transferred estate.29 However, he could not create a servitude on one estate in favour of another estate if he transferred both estates at the same time. 30 Grotius did not explicitly discuss the reservation of a servitude as a way to create servitudes. Perhaps 2.36.6 concerned the reservation of a servitude when transferring ownership of an immovable property, perhaps not. I will come back to that later. Grotius remarked that a right of servitude could not exist if someone were the owner of both a dominant and a servient estate.31 In Roman law, a right of servitude was lost when one person became the owner of both the dominant and servient estate. The Roman jurist Gaius (c. 110-80 ce) remarked: Praedial servitudes are extinguished by merger if the same person becomes owner of both estates.32 Roman jurists used the legal maxim nulli res sua servit (one's own property serves no one) to justify the loss of the servitude.33 The idea seems to have been that an owner already had all powers in the property, and a servitude would have been superfluous.34 The jurists did express this notion with regard to servitudes, but Ulpian wrote the following about a right of usufruct: The only person who can claim at law that he has the right to use and enjoy property is the man who has the usufruct of it. The owner of the estate cannot do so, as a man who has the ownership does not have a separate right of use and enjoyment; the fact is that a man's estate cannot be subject to a servitude in his own favour (…).35 Roman jurists considered a usufruct a personal servitude and put it often on par with real servitudes.36 Therefore, it is likely that Ulpian's explanation for usufruct also applies to the maxim nulli res sua servit.37 Servitudes that were lost by a merger did not revive automatically if the owner of two estates transferred a former dominant or a servient estate to someone else. The jurist Paul advised parties to create a new servitude at the moment of transfer of one of the estates if they desired the recreation of a lost servitude: If a man acquires a house by conveyance after purchasing it, and the house is burdened with a servitude in favour of another house of his, the servitude is extinguished by merger. Should he want to resell the house, the servitude must be expressly recreated, otherwise the house will be sold free of it.38 As mentioned earlier, the transferor could reserve a servitude. The jurists recognised the legal obligation to create a new servitude in cases where a merger was triggered by the law of succession. If the owner of a dominant estate inherited the servient estate, the servitude was lost by merger. In some cases, the heir would be obligated to transfer parts of the inheritance to the beneficiary of a legacy (per damnationem). If the heir needed to transfer an estate that had been encumbered with a servitude in favour of the heir's estate before the death of the testator, the beneficiary had to tolerate the creation of a new servitude.39 The heir was allowed to reserve a servitude when he transferred the former servient estate to the beneficiary.40 On the contrary situation, where the merger had extinguished the servitude, and the beneficiary was entitled to a dominant estate, the heir was required to create a new servitude in his former servient estate.41 Reviving lost servitudes was not only required in cases where an heir transferred former servient or dominant estates, but also in cases where the heir transferred the complete inheritance. Pomponius stated: If I become heir to a man whose estate was servient to mine and I then sell the inheritance to you, the servitude ought to be re-established, to exist as it did formerly, because the understanding is that you are considered as if you actually were the heir of the deceased.42 Re-establishing or reviving the servitude meant that the heir and purchaser of the inheritance needed to create a new servitude.43

The Implied Grant of Servitudes in Roman Law
The main rule was that servitudes needed to be created by express agreement.44 In Dig. 18.1.66 pr (Pomponius 31 ad Quintum Mucium), Pomponius stated: When land is sold, certain obligations are due, even if not stated, such as that the purchaser shall not be evicted from the land or the usufruct of it. Other obligations are due only if made express, such as that the rights of way and of drawing water will be forthcoming; the same is true of urban servitudes.45 According to Paul, it was insufficient for a transferor to state that an estate was encumbered with a servitude generally.46 The express reservation needed to be more specific. Roman law recognised the implied grant of servitudes in specific cases. It concerned cases where the owner of two estates used one of his estates in a servient way in favour of his other estate. Because of the rules on a merger, no servitude could exist. If the owner transferred the actually servient estate to someone else, the transfer sometimes led to the creation of a servitude. A very influential text on this matter during the reception of Roman law was Dig. 33.3.1 (Julianus 1 ex Minicio). Julian discussed a case where the owner of two connected shops bequeathed the shops to two different beneficiaries. One shop supported the other shop. The question was raised if the beneficiary should tolerate his shop supporting the other shop. Julian thought so and added: (..) Let us see whether this is true only if either this servitude is specifically imposed or the legacy is given in the form: 'I give and legate my shop as it now is.'47 According to Julian, the heir was not required to reserve the servitude expressly in his will. The interpretation of the will led to the implied grant of the servitude.48 Roman jurists even recognised the creation of servitudes in cases where nothing had been said on the matter. Ulpian discussed the following case: If the seller of a farm specifies that a burial place should be available for him and his descendants to be buried there, if he is denied access for the purpose of carrying someone for burial, then he can bring an action. For the agreement between the purchaser and the seller is held to guarantee right of way through the farm for the purpose of performing a burial.49 Ulpian's interpretation of the reservation of the burial place meant that the seller had to have the right of way to get there after the transfer of the estate.
Marcellus discussed a similar case in Dig. 8.2.10 (Marcellus 4 digestorum). The owner of two houses (binas aedes) bequeathed one of the houses. The bequeathed house could only be accessed through the house of the heir. The question was put to Marcellus if the heir was required to create a servitude to provide access to the legatee's property. Marcellus answered in the affirmative and stated that the legacy would be useless without access to the property. Interpretation of the legacy meant that the creation of a servitude was required.50 Papinian discussed a case where the owner bequeathed two houses with a common roof to two different beneficiaries.51 According to Papinian, both beneficiaries had to tolerate that the non-common beams of each of the beneficiaries were inserted into each other's walls. This advice suggests the implied grant of a servitude.52 However, there is a different explanation to this advice if we consider this advice in conjunction with a case discussed by Javolen. A testator had bequeathed two houses with a common roof to two different beneficiaries, as in the case discussed by Papinian. Javolen stated that both beneficiaries could not bring actions against each other to prevent each other's beams sticking in the wall, since it was a common wall: '(…) for it is agreed that if a partner owns something jointly, he has all the rights pertaining to it, and so an arbitrator for dividing common property must be appointed in the matter.'53 The rejection of the actions was based on the fact that the parties were co-owner of the wall. They were not entitled to institute such action individually but only together.54 In short, the implied grant of servitudes was not easily accepted in Roman law. A transferor could expressly reserve a servitude when he transferred an estate to someone else. The interpretation of the iusta causa for the transfer could lead to the implied grant of servitudes, especially where servitude was essential to one of the estates.55

Bartolus on Implied Servitudes
Medieval jurists were reluctant to accept the implied grant of servitudes. The glossa ordinaria considered the abovementioned text Dig. 18.1.66 (Pomponius 31 ad Quintum Mucium) in vendendo to be the general rule. If parties wanted to create a servitude, they had to agree so expressly.56 If a servitude was not explicitly mentioned, the jurists were only willing to accept a wide interpretation of legacies and not contracts.57 The outcome of Dig. 8.2.10 (Marcellus 4 digestorum) binas aedes was considered an example of a wide interpretation of a will. In that specific case, the interpretation of a legacy led to the creation of a servitude that was necessary to gain entry to the bequeathed estate.
Bartolus (1313 -57) stated that a seller who wanted to reserve a servitude in the sold property needed to do so expressly in the contract. He wrote: 'Whoever wants to create a servitude must create her expressly.'58 However, he was less reluctant than his contemporaries in accepting implied servitudes if parties had made no mention of any servitudes. Bartolus forwarded his approach in his comments on Dig. 33.3.1 (Julianus 1 ex Minicio) duas tabernas. In this aforementioned text, Julian accepted an implied servitude in the case where the owner of two connected shops bequeathed one 'servient' shop to one beneficiary and one 'dominant' shop to another beneficiary. According to Bartolus If one house sustains a state of servitude which has a continuous and permanent cause and it (i.e. the house) is alienated, (then) the established servitude seems in doubt (…) and I believe the same for the sale of a house, that a servitude seems established, so that it (i.e. the house) bear permanent burdens which it already sustained, especially if the contracting parties knew.
Bartolus's approach is a matter of interpretation of the title (iusta causa) for the transfer of ownership. Parties were free to agree on what happened to actual servitudes after a transfer. If they made no arrangements or if the parties' meaning was unclear, Bartolus was inclined to accept the implied grant of a servitude. According to Bartolus, recognising a servitude was even more obvious in cases where an owner transferred an actually servient estate. Both he and the transferee were aware of this actual servitude (maxime si contrahentes hoc sciebant).
The requirement that the servitude had a continuous and permanent cause protected a transferee, as they entailed some kind of publicity. Examples of servitudes with a continuous and permanent cause were the right to suffer that the beams of a neighbouring estate were inserted into someone else's wall or the right to suffer that a shop supported someone else's shop. Bartolus also mentioned some servitudes which did not meet the requirements, such as a right of way. A right of way was not exercised continuously, but only each time someone walked there. Such servitudes still needed to be created expressly.
Bartolus's interpretation of the implied grant of servitudes was influential in the following centuries. For example, Bartolomeo Cepolla adopted Bartolus's approach in his fifteenth-century monograph on servitudes.60 Two cases brought before the Sacrum Consilium of Naples also echoed Bartolus's approach. In one case, the owner of two houses sold one of them 'free and unencumbered by any servitude' (liberam et expeditam ab omni servitute), except for a specific servitude regarding windows overlooking the garden of the house. The seller made no mention of any other servitudes at the moment of the transfer. After the transfer, the seller claimed there were also implied servitudes regarding other windows and the drainage of water. He argued that these servitudes had a permanent cause, that he and the purchaser had seen these servitudes and they had been aware of them at the moment of the transfer (viderant & sciverant). The purchaser contested the servitudes and argued that only the expressly mentioned servitudes were created. The Sacrum Consilium ruled that only the expressly mentioned servitudes were created, since the contract stipulated that the house and garden were free of servitudes, except for the specifically mentioned servitudes.61 In another case, the Sacrum Consilium discussed how Bartolus left room for the recognition of a servitude, even if the parties were unaware of the servitude. 62 The Consilium considered that a bit harsh, especially if the purchaser would not have paid the same price for the land had he been aware of the servitude. The Consilium recognised the existence of a servitude but decided that the purchaser had paid too much for the land. 63 Simoncelli has pointed out that other contemporary sources required that implied servitudes not only had a continuous and permanent cause but that they needed to be visible too.64

Bartolus's Influence on Early Modern Local Laws
Bartolus's approach has been influential in a large number of countries, including France. 65 The middle and the southern parts of France, the so-called 'pays de droit écrit' , recognised the (written) Roman law as its customary law. Its interpretation by medieval legal scholars, such as Bartolus, influenced the law directly. Bartolus's approach to implied servitudes was adopted, and this meant that the transfer of an actually servient estate led to the implied grant of a servitude if its cause was continuous and permanent.66 This implied grant of servitudes became known as the destination du père de famille (designated use by the head of the family). 67 The written customary law of Paris, the Coutume de Paris of 1510, also recognised the destination du père de famille.68 Since a iusta causa was required for the creation of servitudes, article 91 provided: Destination du père de famille vaut titre (designated use by the head of the family equals title).69 According to Pothier, the rationale behind this provision (in Orléans) was that someone acquired the servient estate 'as is, where is' .70 It remains unclear whether or not this rule applied to all types of servitudes or just servitudes with a continuous and permanent cause.71 From 1580 onwards, a transferor needed to provide written evidence to prove that his former estate had been an actually servient estate. He could provide, for example, a written contract with a contractor, receipts from a construction worker or a deed with a description of the estate before the transfer.72 Bartolus's approach also had an impact in the Low Countries. Antwerp's compilation of customary law of 1582, the so-called Consuetudines impressae contained an elaborate provision on the implied grant of servitudes.73 Title lxii, Van erf-scheydinge, servituten ende des daer aen cleeft, article 14: And so when the owner of two houses or estates sells just one of them or part of it, or with land and its accessories, or just as he had possessed or used it, or with similar words, the watercourses, passages, lavatories, lights (waterloopen, doorgangen, weerdribben, lichtscheppinghen) and other servitudes or benefits, serving both houses respectively, should stay 68  the same as they were before or at the time of the division and sale. And they will be considered to be created tacitly in this case, unless the parties agree otherwise.74 Antwerp's Consuetudines compilatae of 1608 had a similar provision.75 The Antwerp provisions seem to have been the inspiration for the Middelburg Ordinance of December 19th 1617,76 and the Flushing (Vlissingen) Ordinance of March 11th 1628. The latter says: And so when the owner of two houses and estates sells just one of them or part of it, or with land and with all the appurtenances, just as (or with such liberties and servitudes) he had possessed or used it, or with similar clauses and words, the watercourses, passages, lights (Water-loopen, Doorganghen, Licht-scheppinghen) and other such uses or benefits should stay, and shall be used, in the same way as they were before and at the time of the division and sale. And they will be considered to be created tacitly in this case, unless the parties agree otherwise. The same applies in the case where two houses are divided, simply, or with the aforementioned words, between the heirs or successors of the possessor.77 The approach taken by the cities of Antwerp, Middelburg, and Vlissingen meant that a purchaser could acquire an encumbered estate even if he was unaware of a servitude. Just as in Bartolus's approach, this was a matter of interpretation. Interpretation of the clause 'just as the seller had possessed or 74 14. 'Ende so wanneer de Proprietaris van twee huysen oft erven een van  used it' led to the purchaser acquiring an encumbered estate. The parties could agree that no implied servitudes would be created.

Interpreting Grotius's Position on Implied Servitudes
Grotius's Introduction to Dutch Jurisprudence was published in 1631, three years after the introduction of the provisions on implied servitudes in the local laws of Vlissingen (1628).78 The Introduction 'is a work of masterly systematisation, a condensed summary of existing laws (…)' , as R.W. Lee put it in 1930.79 In the paragraph on the obligations of a seller, Grotius noted that a seller is bound to deliver a property free from servitudes (vrij van alle dienstbaerheden):80 Under warranting is also understood freeing: for the seller must deliver to the purchaser the thing sold free of all servitudes, or otherwise he is bound to make compensation as far as the purchaser is affected thereby, the purchase still remaining intact (…).81 wanneer een eyghenaer van twee Huysen ende Erven, een van dien, ofte een deel daar af vercoopt simpelyck, ofte met zyn ghevolgd ende toe-behoorten, ghelyck (ofte met sulcke vryheden ende servituyten) een sulke dat beseten ende ghebruyckt heeft, ofte diergelyke clausulen ende woorden, moeten de Water-loopen, Doorganghen, Licht-scheppinghen, ende andere dierghelycke ghebruycken ofte commoditeyten blyven, ende voorts gebruyckt werden, gelycke voor ende ten tyde vande splissinge ende vercoopinge geweest hebben, ende worden in sulcken cas verstaen de voorsz: servituyten stilswyghende gheconstitueert te zyn, ten ware daer in anders by partye ware voorsien, ende sal 't gunt voorsz: is mede plaetse hebben in twee huysen, die by den erf-genamen ofte successeurs van  that 'the benefit of the servitude was only suspended for as long as that time had lasted' .86 Several different early modern scholars have discussed Grotius's opinion on servitudes as a matter of interpretation of the iusta causa. 87 Voet rejected Grotius's opinion and considered to be on par with Bartolus's approach on implied servitudes. He wrote 'confer' after Grotius's opinion and referred to the aforementioned case of the Sacrum Consilium of Naples and treatise of Bartolomeo Cepolla ( §5).88 Unfortunately, he did not elaborate on the extent of the similarity between Bartolus's approach and that of Grotius's. Voet explained that the Roman jurists were only willing to accept the express reservation of a servitude by the seller, or the implied reservation, if a seller had sold his properties 'as they are' at the time of the sale. 89 Voet adhered to the general Roman rule for the creation of servitudes and seemed reluctant to accept the various exceptions to the general rule that early modern scholars had distilled from Roman law. Voet's disagreement with Grotius seems to have been unconnected to their disagreement over whether the creation of a servitude required a quasi-delivery before a local court.90 At no point did Voet refer to how Grotius's approach (or his own) related to the required quasi-delivery.
At the end of the eighteenth century, Didericus Lulius, Pieter and Reinier van Spaan, and Joannes van der Linden briefly discussed Grotius's approach. They preferred his opinion over Voet's as they argued that Grotius's opinion was corroborated by the local laws of the Middelburg, Vlissingen, Schoonhoven, and Dordrecht.91 In these cities, the interpretation of the clause 'just as the seller had possessed or used it' led to the purchaser acquiring an encumbered estate.92 Apparently, these scholars believed that Grotius's opinion was a matter of interpretation of the iusta causa and that the creation was not restricted to the revival of lost servitudes.93 They did not, however, elaborate on the meaning or origin of Grotius's position.
In the South African case, Salmon v Lamb's Executor & Naidoo (1906), judge Kotzé paid considerable attention to the interpretation of Grotius.94 Kotzé remarked that the passage 'seems capable of a twofold meaning' .95 The first meaning is that Grotius referred 'to the case of a person who as owner has built two adjoining houses or has purchased them, and during his ownership has so arranged that the one home is made subject to a burden in favour of the other.'96 The second and, according to Kotzé, the 'generally-received interpretation' , is that the owner of a dominant or servient estate 'acquired the other, whether servient or dominant and so continued to enjoy the benefit of the privilege of servitude which the one house previously possessed over the other' .97 The subsequent transfer of one of the estates then led to the revival of the old servitude. Kotzé then mentioned the legal scholars who adopted this second interpretation, such as Wassenaar, Van Streyen, and Van Leeuwen.98 Kotzé remarked: 'As I have already stated, Grotius throughout his Introduction does not refer to any authorities in support of his text, and no commentator who has followed him has cited any general law or general custom, nor any decision of the Dutch courts in support of his statement.'99 Kotzé dismissed the meaning of the abovementioned local laws that corroborated Grotius's opinion as he deemed them 'purely local in their nature and operation' and not 'declaratory of the general law of the land.'100 Kotzé remarked: Had the authors of the Rechtsgeleerde Observatien cited a general statute or placaat, or a decision of either the Court or the Supreme Court of Holland, agreeing with the statement of Grotius, or a general Dutch custom to the same effect, the case would have been very different. (…) As there exists no general law or custom, nor any decision of the Dutch courts to 93 Lulius Kotzé reluctantly concluded that Voet's position was the correct statement of Dutch law and held that the creation of servitudes in South Africa required an express grant. Unfortunately, Kotzé overlooked Voet's observation that Grotius's position was comparable to Bartolus's and the fact that the abovementioned local laws progressed from Bartolus's approach to implied servitudes. Since Bartolus's approach is an interpretation of Roman law, it can be argued that these local laws did not depart from Roman law. In my opinion, Voet's observation that Grotius's position was comparable to Bartolus's was right, and I will now elaborate on this. During the years that Grotius worked on his Introduction, he had access to, amongst other things, Bartolus's commentaries on the Digest and the published customary laws of Antwerp.102 Looking at Grotius's position on servitudes, it is hard to avoid the impression that Grotius had progressed from Bartolus's approach to implied servitudes, despite the fact that Grotius did not mention whether or not it was a matter of interpretation or that the servitude needed to have a continuous and permanent cause, or that it had to be visible.
Grotius's choice of words suggests he was familiar with the learned law approach to implied servitudes. The key fragment on implied servitudes, Dig. 8.2.10 (Marcellus 4 digestorum), started with the words 'two houses (binas aedes)' .103 Medieval comments on this text stressed that parties needed to create servitudes expressly and that there was some room for the wider interpretation of legacies. Grotius's wording also echoed Bartolus's approach. In the key text discussed by Bartolus, Julian mentioned 'two shops (duas tabernas)' and the legacy of one of them.104 Grotius also discussed two houses and the transfer of one of them. Furthermore, Grotius's wording is almost identical to the provisions on implied servitudes in Antwerp's Consuetudines compilatae of 1608 and the Middelburg Ordinance of December 19th 1617 ( §7).
There is another reason why it is not very likely that Grotius's opinion was restricted to the revival of former servitudes, as Van  Keessel have argued. Roman jurists often discussed the revival of servitudes in one breath with the extinction of servitudes. Grotius did not discuss the implied grant of servitudes within the context of a merger. He inserted it into the paragraph on the creation of servitudes. Grotius did not mention whether or not his opinion on servitudes was a matter of interpretation. He even added the words, 'without any mention either the one way or the other ('zonder iets uit ofte in te zeggen')' , which suggest that the parties' silence sufficed. In one of his other books, De iure belli ac pacis libri tres, Grotius paid quite a bit of attention to interpretation.105 The chapter de Interpretatione concerned the interpretation of laws and treaties, but many parts are equally applicable to the interpretation of contracts between two private parties. Grotius wrote: Where we have no other conjecture to guide us, words are not to be strictly taken in their original or grammatical sense, but in their common acceptation, for it is the arbitrary will of custom, which directs the laws and rules of speech.106

Grotius's Influence on the Implied Grant of Servitudes in the Netherlands
After the publication of Grotius's Inleiding, several different cities expressly adopted the implied grant of servitudes in their published local laws, such as Rotterdam (1654), Schoonhoven (1644), and Dordrecht (1725).107 These cities did, however, expressly consider the implied creation to be the result of the interpretation of specific clauses in the contract. This means that these provisions are more similar in wording to the local provisions of Antwerp (1582), Middelburg (1617), and Flushing (1628) than to that of Grotius's Inleiding (1631). Grotius may have paved the way for Rotterdam, Schoonhoven, and Dordrecht, but this influence on these local provisions cannot be proven.
Apart from these cities, some reluctance to accept the implied grant of servitudes remained. The well-known collection of legal opinions entitled the Consultatien, Advysen en Advertissementen gegeven ende geschreven by verscheijden Treffelijke Rechtsgeleerden in Hollant en elders (commonly known as the Hollandsche Consultatien), provides an example of this reluctance.
In 1644, a purchaser of a house contested the implied grant of a servitude.108 Before the transfer, the seller had been the owner of two neighbouring houses. One of the houses discharged rainwater from its roof onto the roof of the other house. The seller sold the 'servient' house with the clause 'that the sold house would be accepted with such servitudes as it was encumbered with' .109 After the transfer, the seller, who was still the owner of the 'dominant' house, argued that a servitude for the drainage of rainwater and run-off had been created. In his opinion of this case, the jurist Moons dismissed the position of the seller. Moons argued that the owner of two houses could not effectively create a servitude in one house in favour of the other house. According to Moons, the aforementioned clause only constituted an exoneration in the sense that the seller was not liable for any servitudes with which the house may have been encumbered. It did not qualify as the reservation of a servitude. Moons emphasized that the reservation of a servitude needed to be made 'expressly and specifically' .110 The Supreme Court of Holland, Zeeland, and West-Friesland was less reluctant to accept the implied grant of servitudes. In 1709, it had to decide whether a contractual clause entailed the implied reservation of a servitude. Earlier, in 1641, the owner of the dominant estate had acquired ownership of the servient estate. After thirty years, he sold both estates to two different purchasers. The seller sold the former servient estate 'on all the conditions, on which he had purchased it himself (op al de conditiën, waarop hij het zelf had gekocht)' . Subsequently, the purchaser blocked the owner of the other estate from exercising his servitude (water loop).
The Supreme Court formulated the general rule that a servitude that was lost by confusio did not automatically revive after one of the estates was transferred.111 The seller needed to reserve a servitude, if the parties desired the revival of a lost servitude. In his private account of the case, Justice Cornelis  112 The Supreme Court ruled that the interpretation of the cited clause implied the reservation of a servitude. Unlike in France, the creation of servitudes by means of destinatione patris familias never really took off in the Dutch Republic. One of the reasons seems to be that the most prominent advocate of this approach in the Republic, Grotius, could have been more clear about the meaning of his position in the Inleiding to avoid confusion and misinterpretation.113 Moreover, I believe that Dutch legal practice was less inclined to recognise the implied grant because it was at odds with the publicity of real rights. 114 In the Dutch Republic, a fairly adequate system of public registration of the transfer and encumbrances of immovable assets existed. In 1560, the Spanish king Phillip ii provided that all transfers and encumbrances regarding immovable assets be registered in the records of the local courts. He noted that abuse and inconveniences occurred with regard to the sale and encumbrances because of a lack of adequate registration. The required registration meant that a secretary of the Court transcribed (a part of) the contents of the deed of a pledge in a publicly accessible register or record. The States of Holland maintained these rules on registration in the Politieke Ordonnantie of 1580.115 Publicity enabled prospective buyers to establish a person's entitlement to the immovable assets and credit providers to assess a person's creditworthiness. 116 Interested parties had to look in multiple registers to establish whether or not assets were encumbered. However, they could also make use of the (paid) services offered by the city's secretary to look up any encumbrances.117 Recent research suggests that the registers were consulted frequently.118 If servitudes could be created by means of an implied grant, third parties could be unaware of such an encumbrance when they inspected the public register. In the Kingdom of France, on the other hand, publicity of transfers and encumbrances of immovable assets was lacking. A lack of public registration of (implied) servitudes was less problematic since third parties were unable to inspect public records for transfers and encumbrances anyway. The requirement that the implied servitudes needed to have a continuous and permanent cause was an alternative way to achieve publicity.
The implied grant of servitudes was eventually adopted in all of Holland (and the rest of the Netherlands). In 1810, the French emperor Napoleon Bonaparte annexed the Kingdom of Holland, the successor of the Batavian (and before that Dutch) Republic. In 1811, he introduced the French civil code of 1804, and it combined the various customary French approaches to implied servitudes. Article 692 provided: the designated use by the head of the family equals title for continuous and apparent servitudes.119 If a seller transferred an estate with a visible sign of servitude ('un signe apparent de servitude'), a servitude was created, even if the parties had made no mention of a servitude.120 After the defeat of Napoleon, the Low Countries regained their independence and strived to introduce their own civil code. The Dutch civil code of 1838 contained a nearly literal translation of the French provisions on implied servitudes.121 Neither Bartolus's nor Grotius's position on implied servitudes was discussed in the creation of the civil code. The French way to create implied servitudes, however, was never fully embraced by legal scholars in the Netherlands because it was at odds with publicity.122 This explains why it was not included in the new Dutch civil code of 1992.123 Again, neither Bartolus's nor Grotius's position on implied servitudes was discussed in the creation of this civil code.

Conclusion
One of the book chests in which Hugo de Groot escaped from Loevestein castle was previously used to deliver the sources that Hugo de Groot used for writing the Inleiding tot de Hollandsche Rechts-geleerdheid. Bartolus's commentaries on the Digest, the published customary laws of Antwerp, and various other legal sources enabled Grotius to formulate his position on implied servitudes. According to Grotius (2.36.6), the actual use of two houses by the same owner could lead to the implied grant of a servitude, if he transferred one of the houses to someone else, 'without any mention either the one way or the other' .
The 'generally-received interpretation'124 that the implied grant was restricted to the revival of servitudes that were lost by confusio, is unfounded. However, as Voet suggested, Grotius most likely progressed from Bartolus's approach to implied servitudes and the local laws in the Low Countries that it influenced. According to Bartolus, a servitude was implied if there was doubt about the meaning of the parties and the servitude had a continuous and permanent cause (causam continuam et permanentem). The local laws of Antwerp and Middelburg expressly recognised the implied grant of servitudes that had a continuous and permanent cause, such as watercourses and passages. Grotius had access to these sources, and his choice of words echoed these local laws and the key texts from the Digest on implied servitudes that were discussed by Bartolus. Grotius did not mention whether the implied grant was restricted to servitudes that had a continuous and permanent cause. However, as the Inleiding was a condensed summary of existing laws, it is unlikely that Grotius formulated a new rule for the implied creation of servitudes that deviated from the ius commune and customary law.
Unlike in France (and many other European countries), the creation of servitudes by means of destinatione patris familias never really took off in the Dutch Republic. One of the reasons seems to be that Grotius, as the most prominent advocate of this approach in the Republic, could have been more clear about the meaning of his position in the Inleiding to avoid confusion and misinterpretation. Moreover, Dutch legal practice and scholarship were less inclined to recognise the implied grant because it was at odds with the publicity of real rights. In the Dutch Republic, a fairly adequate system of public registration of the transfer and encumbrances of immovable assets existed. If servitudes could be created by means of an implied grant, third parties could 124 Salmon be unaware of such an encumbrance when they inspected the public register. Even after the French code civil was introduced, the implied servitude was never fully embraced by legal scholars in the Netherlands because it was at odds with publicity. This explains why it was not included in the new Dutch civil code of 1992.