STOP Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact support@jstor.org. THE "UNINUNDATED LANDS" IN PTOLEMAIC AND ROMAN EGYPT By W. L. Westbrmann PART II ATTITUDE OF THE GOVERNMENT, OWNER, AND TENANT The &/3poxos yrj in Egypt was that land which was not reached by the Nile flood, but was actually capable of irrigation by the expendi- ture of extra labor in running deep field laterals into it. The amount of the &@poxos in any section varied with the height of the annual inundation. The important fact as to the government's attitude toward this category of grain land is that, in so far as our present testimony goes, the rulers of Egypt demanded and received from the "unflooded" land certainly as high a rate of taxes and rentals as it received from the flooded lands, possibly even a higher rate. The proof of this statement is to be found in an analysis of P. Brux. I. Previous discussions of this important document have always been conducted upon the assumption that the administrative, or ownership, categories (temple land, royal domain, private land) were the all-important factors. Actually the vital distinctions are the production categories (flooded, unflooded, and dry). Of P. Brux. I there has been preserved the lower part of eleven columns, about two-thirds of each column being lost. 1 The outer portions of columns I and XI are gone. The tabulation in this record is by "sown land," SpaYcoY6s) runs along the entire length of its western boundary, a main ditch (5uopt;£) along a part of the northern boundary, and another supply ditch forms a part of the southern boundary. Other agencies than the difficulty of artificial irrigation must have weakened the productive character of the soil of the unflooded tracts in this division. Taking the average of the four divisions together 1 the flooded land pays at 1 .285+, the unflooded at 1 .355+artdbae to the aroura. The government therefore receives as an average 7 per cent more from the unflooded land in these divisions than from the flooded. II. RENTS FROM j&wnAi^ yfj, ROYAL DOMAIN 7th Division Arourae Paying Average per aroura 139 artabae wheat I Pays at 4£ artabae 1290 artabae wheat j Pays at 4§+ artabae 8th Division (No flooded land) Unflooded land . . ..I 280 I 1319 artabae wheat I Pays at 4rV artabae I I (plus a little barley) j As revised by the inspection (iiriaicexl/K) Unflooded land .... I 101 j 408 artabae wheat I Pays at 4 artabae j ) (plus a little barley) | 1 The average is taken from the sums given in the estimates before the inspection and the consequent revision. The figures given in the revisions of divisions 8 and 9 would not, however, materially affect the result. 2 In the four divisions of this village district upon which the information has been preserved, it is noteworthy that most of the land easily reached by the inunda- tion was held in private ownership. Flooded Land 2 I 31 Unflooded Land.... 283 172 W. L. Westermann Unflooded land . Unflooded land . Unflooded land . . . 9th Division (No flooded land) 63 I 251 artabae wheat (plus ! 68 artabae barley) As revised by the inspection 41 I 166 1 artabae wheat ' (plus 3J art. barley) j 24 10th Division (No flooded land) 120 artabae wheat (plus 11 barley) Pays at 4 in wheat (plus 1 in barley) Pays at 4 artabae Pays at 5 artabae wheat (plus J— barley) According to this estimate flooded land was expected upon the royal domain only in the seventh division. Consequently it is only there that we have the possibility of comparing the rents upon the flooded and the unflooded lands. 2 The result is the same as that obtained in dealing with the taxes upon the private lands. Again the unflooded pays a slight increase over the flooded land. This conclusion is entirely supported by B.G.U. I, 84, of 242/43 a.d., found in the Fayum. It is the account of a village scribe, whose name is lost, and reads (11. 4 ff.) : .... for the exaction of grain revenues of the exchequer and the patrimonial estates 3 through the state peasants for the sixth year of our Lord Imperator Caesar Marcus Antonius Gordianus Pius Felix Augustus. The sum total is : From Pelusium: 4 Due the exchequer and the patrimonial estates through the state peasants: Arourae [5] 5 517f \ A", paying in wheat artabae 29.299H A- ~h, in barley artabae 302§|, in beans 1261f fr -&. Of this 1 In Mus&e Beige, VIII, 114, a printer's error gives the resolution of p{C (p. 108) as 107J. 2 The inspection showed flooded fields (ffePpeynivri) in the eighth and ninth divi- sions. But as their payments are not noted down by the inspectors we cannot use them. 3 els kiraiTTiaiv 6po)v 5totic^}> 5ta 5^(juopy[av lr «]<£if e t){5, which is my own, as certain. The substitution of 4 or 6 would in either case give an impossible payment of rentals upon state lands which are inundated. For explanation of the fractions see Wilcken's notes to B.G.U. 84. 'Uninundated Lands," Ptolemaic and Roman Egypt 173 amount the unflooded land (o/?poxos) is arourae 4,437f J -fa paying in wheat artabae 24,565 -fa , in barley artahae 2975 i> in beans The remainder is arourae Tabulation of this account, again disregarding the fractions, gives the following result: RENTS UPON STATE LANDS Total. Unflooded land . Arourae 5,517 4,437 Paying 29,299 artabae wheat 302 artahae barley 1,261 artabae beans 24,565 artabae wheat 297 artabae barley .... artabae beans Average per aroma Pays at 5J+ (Negligible) (Negligible) Subtracting the amounts of unflooded from the total we have: Flooded . Arourae 1,080 Paying 4,734 artabae wheat 5 artabae barley Average per aroura Pays at 4J— (Negligible) Again the result is quite clear. The unflooded land, granting that my restoration of the total arourae to 5,517 is correct, here paid a much higher rent than the flooded, in fact a full 25 per cent increase. In Menches' report of 114/13 b.c. there appears in the list of the lands producing no revenue for the state 39| arourae once "unflooded" (a/Jpoxos), 1 which in the year 114/13 B.C. were classified as tup iv eirio-Tacret,. This phrase, which still lacks a satisfactory explanation, 2 need not confuse one in regard to this bit of " unflooded " land. The essential fact is that these 39| arourae so long as they were still productive in the "unflooded" class, paid rent at 4| § T V artabae to the aroura. A basis of comparison of this rent on unflooded land with the state's income from flooded lands is found in P. Teb. I, 71, where Menches gives the statement of the area of Kerkeosiris for the year 113 b.c. For the "flooded and sown land" 3 the area was 1193§ | arourae, the rental of which was 4665J rV artabae. The average rent per aroura on the flooded land was therefore only 1 P. Teb. I, 72, 11. 427-39. 2 See the discussion in P. Teb., Appendix I, pp. 576-80. I have not been able to find its technical meaning in any of the papyri published since those in P. Teb. I. 3 Lines 2-3, TrpootvyyeKna. tj)s fltfipey (terns ml ka"7rapiikvr\s. 174 W. L. Westekmann 3-fTs+artabae, which is considerably lower than that exacted from the 39§ arourae of unflooded land cited above. The fact about the unflooded land, in so far as our present data go, is that it pays on a different system from the flooded land, and quite as much if not more than the flooded. In other words the Egyptian government was attempting to force production on the unflooded land, to compel the owner or leaseholder to see to it that the extra labor required to irrigate the unflooded land should be undertaken. 1 The whip used to enforce this was a high tax or a high rent upon the unflooded land. It is the principle lying at the base of the single-tax theory. A similar measure, though the circum- stances differ somewhat, was proposed by Lord Grey in 1833 at the time of the Emancipation Act of the Jamaica negroes. 2 The negroes derived their provisions, under the plantation slave system, from their own provision grounds. Foreseeing that they would not work on the plantations for wages, but would work only enough to produce a living from these lands, Lord Grey proposed a tax on the provision grounds. The Jamaica legislature, however, refused to adopt the measure. The system in Egypt was clearly so devised that there would be no escape from the taxes in kind upon the a^poxos yrj. The owner or the lessee declared the amount of land which had been left unflooded by the previous inundation, and was, therefore, "unflooded during the present year." 3 Those declarations upon which the date is preserved, fall in Mecheir (P. Oxy. VIII, 1113); Phamenoth (P. Oxy. XII, 1459, 1549); Pharmouthi (P.S.I. Ill, 161, P. Teb. II, 324); and Epiph (P. Fay. 33), that is, approximately from the beginning of February to the middle of July, which is the period of low Nile just preceding the flood time. The amount of the unflooded thus declared for any given year was used as the basis 1 Similar in its purpose and character is the insistence that the rent must be exacted upon flooded land which, by neglect of the cultivators, has not been sown. P. Teb. I, 67, 1. 71. 2 Egerton, Hugh E., A Short History of British Colonial Policy, London, 1897, p. 330. Lord Grey was the British Colonial Secretary from 1846 to 1852. 3 P. Teb. II, 324; B.G.U., I, 139, ij/3/3ox'/'"''Os ; P. Grenf. II, 56, 1. 10, fi0poxv^"at. The translations of the editors in P. Teb. 324 ("as having been unirrigated") and in P. Grenf. II, 56 ("had not been irrigated") are very misleading. "Uninundated Lands," Ptolemaic and Roman Egypt 175 for the inspection of the following year, which took place after the inundation. This explanation makes it possible to understand more clearly that part of the priestly praises of Ptolemy III in the Canopus inscription which deals with the low Nile and its results. 1 A particular year of low Nile is mentioned. All the people of Egypt were in terror at what had befallen, recalling the disaster which had occurred under certain of the former kings in whose time it had happened that the natives met with low inundations (afipoxlais here seems to be used in the pregnant sense that they faced wide stretches of uninundated land, aftpoxos yfj). 2 But Ptolemy, taking many precautionary measures and remitting not a few of the revenues for the sake of the safety of his subjects, had grain imported into the country at higher prices, from Syria and Phoenicia and Cyprus and other places; and so he saved the inhabitants of Egypt. The benevolence of Ptolemy III lay in the fact that he sacrificed the interests of the treasury by not exacting the customary high taxes or rents on the unflooded lands, which were in that year so great that the peasants looked with terror upon the prospect of being forced to cultivate and pay in kind upon them all. The context 3 shows that the appoxicu had to do with the grain revenues of the state. It is clear that if the unflooded land had received remission or lowering of taxes in case of low Niles (causing hypoxias), the people would not have suffered to the same degree as the treasury did. But the Canopus inscription shows that the people, and not the state, actually did suffer because of the afipoxos yfj, due to the taxes and rents upon this type of land. The edict of Tiberius Julius Alexander, the Alexandrian Jew who was prefect of Egypt in 66-69 a.d., contains a section 4 which throws 1 Dittenberger, Orientis Graeci Inscr., 56, 11. 14-19. 2 The demotic translation at this point does not serve to clarify the Greek original. E. Revillout, Chrestomathie De"motique, Paris, 1880, p. 133, translates the demotic text as follows: "ils se lamentaient a cause de les choses advenues quand ils se repor- taient aux malheurs arrives (etant) sous les rois qui furent auparavant que il arriva a les hommes qui en Egypte (d') §tre en disette ( ? ?) sous eux." Evidently the demotic text has no word corresponding to d/Spox«us of the Greek text. 3 Especially the airov lAeraTen'p&iJLepoi els ri\v x&pav rtfuov tiei^dvon' of 11. 17—18. 4 Dittenberger, Orientis Graeci Inscriptiones, 669, par. 12. 176 W. L. Westermann some additional light upon the tax collection upon the flooded and unflooded lands. After speaking of the peculations of the eclogistae (the assessors) and the strategi, the prefect proceeds: A similar type of falsification is the so-called "collection based on an average," 1 not based on the current Nile rise but on a comparison of certain former Nile rises. And yet nothing seems more just than the truth. There- fore I bid the people to live confidently and till the soil zealously, knowing that the tax collection will be based upon the truth as to the current Nile rise and the flooded 2 land, and not upon the false record of those who make entries "based on an average." Since the tax estimate and the collection upon the flooded land (jjiefipeyniv'q yrj), as I have shown above, was regarded as "normal" or "uncontested" (6ju6Aoyos), the variability of the taxes as implied in the edict of Julius Alexander must have lain in the amount of the unflooded and the assessment upon it. If my understanding of the unflooded land and its taxes and rents is correct, the injustice worked upon the taxpayer or state lessee by the collection on the basis of an average of several years lay in two facts: 1. The average of the unflooded (aftpoxos) of any given number of previous years might be more than the actual area of the unflooded land in the current year. The peasant might therefore be forced to pay the higher tax on a part of his land, as unflooded, which was actually flooded and subject only to the normal tax (as being " uncon- tested," ojuoXcyos). 2. The entries of the tax officials, eclogistae and strategi, when the estimate and collection were made upon the computation of an average of several years, could not be controlled by the peasant taxpayer or lessee, as they could when the actual amount of the flooded and unflooded land was the basis of calculation. The real amounts of the flooded and unflooded were facts which the peasant knew. Against overexactions and falsified statements he might then make his appeal for redress. The statement of the astute Jewish prefect shows that the truth as to the amount of the flooded land did not worry the taxpayer or the state lessee. It must therefore be the untruth about the 1 fiXfyo/ikpri nari. oifriv iwairrjais, 1. 55. 2 7rpAs tA 4Xij04s tijs ovarii 6.va0/ureai) of the a/3poxos yfj by the landowner. The conclusion thus wrongly drawn was strongly sup- ported by false interpretations of two other types of documents, namely, (1) the many private land leases which we have; (2) con- tracts of Egyptian peasants with the government to lease, at greatly reduced rentals, lands deteriorating in production, until these should again be brought up to a state of high productivity. We have several examples of the latter type of document. They fall in the time of Hadrian when the question of farm abandonment and decreased grain production had become very serious. The facts in regard to the attitude of the tenant, the actual tiller of the soil, toward the afipoxos yfj, as expressed by his demands in the leases which he signed, must be closely considered. This requires a statement of one's attitude upon certain moot points in regard to the incidence of and responsibility for the land tax and land rent in Egypt. Preisigke, 1 Eger, 2 and Grenfell and Hunt 3 hold to the theory that the responsibility for the land tax fell primarily upon the lessee 1 Fr. Preisigke in P. Strasab., I, 23, p. 89, and in Girowesen im griechischen Aegypten, Strassburg, 1910, p. 78. i Eger, Otto, Zum aegyptischen Grundbuchwesen in rijmischer Zeit, Leipzig, 1909, pp. 188-89. 3 P. Fay. 33, note to 1. 18; P. Oxy. II, 268, note to I. 18. Cf. note to 1. 11 of P. Oxy. XII, 1460. 178 W. L. Westermann rather than the landowner. Ulrich Wilcken 1 and Otto Waszynski 2 hold to the view that it was the direct lessee in the case of state land, and the landowner in the case of private land, who was subject to and ultimately responsible for the rent or taxes. General considerations of taxation methods lead to the assump- tion that the ultimate tax obligation rested legally upon the land- owner, or direct lessee of government land, rather than upon the tenant. It is the general rule in empirical taxation. It is also the fundamental legal theory of Roman land taxation. 3 Specifically, for Egyptian land taxation the primary tax obliga- tion of the landowner as against the lessee, or the rental obligation of the direct lessee as against the sublessee, is proven by a number of considerations : 1. In the Ptolemaic land lists kept by Menches, the village scribe of Kerkeosiris, 119-112 B.C., the information booked upon the land register was: the name of the cleruch who held the land as yrj kv air6s. b The important person to the state was evidently the landowner or direct lessee. If the government had held the tenant responsible the name of the land- owner or direct lessee would not have been needed in the record office. 2. In the airoypa4>ai of unflooded land it is always the landowner who makes the declaration, giving the name of the tenants. 6 3. A series of documents of the year 223-22 b.c. 7 gives clear proof of the government's view that the responsibility for the rentals upon its lands rested primarily upon the direct lessor. Three 1 Papyruskunde, I, 1, p. 180. i Die Bodenpacht, Leipzig, 1905, p. 115, n. 1, and pp. 116-17. 3 Bernhard Matthias, Die rOmische Grundsteuer, Erlangen, 1882, p. 61, where the ancient sources are available. 4 P. Teb. I, 62, 11. 180-236 et passim; 63 passim. » P. Teb. I, 62, 11. 195, 199, 203, etc. •P. Oxy. XII, 1459 and 1549; VIII, 1113, col. 1; B.G.U. I, 198; P. Fay. 33; P. Ham. 11. 7 P. Eleph. 15-19. No. 16 is in demotic. "Unintjndated Lands," Ptolemaic and Roman Egypt 179 brothers, Pinyris, Berenebthis, and Psentaes the Little, had held a lease upon 30 arourae of land upon which they paid rent in money. These three were middlemen, as the land was really cultivated by other peasants. 1 The three brothers, after paying one of the four instalments of the rent, became insolvent. 2 In their insolvency statement they had ceded the land to one Xenon. 3 It was neces- sary, however, that Xenon acquire the land by the legal method of submitting to the government officials an application for a lease upon the 30 arourae. This he did, 4 offering to make the remaining three payments of the rent, which amounted to 240 drachmas. 6 This indication that the government held the direct lessee responsible for the rent receives further support from an incident which occurred after the period of the insolvency of the three brothers and before the question of the reassignment of a new lease upon the land had come up. After the insolvency of the three the land continued to be worked by the tenants, i.e., the actual cultivators or sublessees. In order to secure themselves in their occupancy, they submitted an engagement (i>7r6 (11. 24-26). P. Oxy. Ill, 501, 187 a.d. Lease for four years of 1\ arourae at a fixed rent per annum of eight artabae of wheat and forty drachmas. iav Si ris Tois H-rjs Vr&Ji afipoxos wapaSexOy&eTat, t$ nefxurdwuevG). The taxes rest upon the landlords (11. 27-31). P. Oxy. VI, 910, 197 a.d. Lease for four years of five arourae, to be planted in the first and third years in wheat, paying rent at 6 artabae of wheat per aroura; in the second and fourth years to be planted in resting crops, paying at 32 drachmas per aroura. The taxes rest upon the landlord, eav Si rts airo rod bnbvros 'irovs, 8 p.77 eify, afipoxos yivrjTai. irapaSexOrjo-erai tQ fieiuaQupivop. 1 P. Teb. I, 105, 1. 18. ! Ibid., 11. 39-42. 3 Ibid., 11. 48-49. 182 W. L. Westermann The similarity in phraseology in all these documents is notable, implying a constant and customary clause regarding the afipoxos yrj in the leases. Equally notable is the direct connection of this clause with that which places the paying of taxes upon the landlord. The tenant gathers the harvest and takes it to the threshing floor. Here the taxes on the land, for which the landowner is responsible, are exacted, that for the afipoxos as well as that for the PefipeypAvrj yrj. In these leases the tenant has been able to protect himself against the arduous labor necessary to cultivate the afipoxos yrj, by a clause which enables him to deduct the tax on the unflooded land out of the rent, when he settles with the landlord for the year. A different method of reaching the same result is exemplified by P. Strassb. 10, 268 a.d. Lease for one year of 11 arourae of land, a part of which is afipoxos. The flooded land only is to be sown and the rent is upon it alone; 11. 9 ff . : apovpas evhtna, h als afipoxos, els airopav irvpov nal naredecnp xopTaap.aT<>)v Kara rb r)pj,crv rr)v \eip.vaadel- crav yrjv, (fropov luacrT-qs apovpas rrjs iv cnropq Ik yeasperpias (pavrj- crop.ivr)s, etc. This method of avoiding the laborious cultivation of the afipoxos is found also in P. Flor. Ill, 281, of 517 a.d. In the Byzantine period the tenant was unable to shift upon the landowner the entire burden of the tax upon the unflooded land. Those private leases, which we have from this period, are of the sixth century. The rent paid upon the unflooded land is either one-half of that paid upon the flooded land, 1 a reduction upon the rent of the flooded portion, 2 or a fixed rent in money or kind. 3 Characteristic of the attitude of the contracting parties in most of these leases is the phrase implying that the "unflooded" condition of the land is highly undesirable. This feeling is usually expressed in the form of a wish that it may not occur, iav Si tis, 8 pr) city,* afipoxos yeprrrai, or in closely similar phraseology. The failure of the inundation to saturate the subsoil and fructify any part of a farm, works harm to both parties, to the tenant by cutting down 1 P. Grenf. I, 56, 11. 12-33, date 536 a.d. » P. Grenf. I, 57, 11. 10-11, date 561 a.d. 3 P. Lond. Ill, 1006, p. 261; P. Flor. Ill, 286, date 552 a.d. Whether rental upon the fiftjoxos is high or low cannot be determined in these cases. 4 P. Oxy. VI, 910; P. Grenf. I, 56 and 57; P. Flor. Ill, 286. " Unintjndated Lands," Ptolemaic and Roman Egypt 183 the amount of land in one plot which he may readily cultivate, and hence the amount of his income, even though he arrange matters so as to avoid paying rent upon the afipoxos. The landlord is harder hit. His rent is decreased — yet he must pay his high tax to the state upon the "unflooded" land. the afipoxos Y?7 in leases of public land The extant evidence upon the attitude of the lessee of state land toward the unflooded portions (aj3po%os) of that land is equally deci- sive. In these leases the direct lessee, the ormbaios yewpyds, is usually also the tenant or actual cultivator, yeupyds avros. The clear examples are : B.G.U. II, 640. First century. Offer, addressed to the royal scribe, to lease shore land of the state's domain. Amount of rent is lost. If any of the land becomes of the classifications " unflooded " or "waterlogged," a deduction will be made to the lessees from the rent. 1 B.G.U. II, 571. 2 Script of the second century. Contains in lines 8-11 a summary of a lease of one aroura of unproductive "dry" land (xeperos) to a certain Harpocration at 2 drachmas. The lease contained a supplementary notation that the tax was not to be exacted, and one regarding the deduction of the "unflooded and waterlogged land," d/3p6xou km Ka6v8(arov). 3 B.G.U. Ill, 831, 201 a.d. Offer to lease eighteen arourae of " dry " land belonging to the state. If any part becomes " unflooded " or "waterlogged" the rent will be reduced, lines 14-16: &d.v Si tis o/3poxos rj Ka.6v8a.TOS yeyrjTat., irapabtx^W^ M 01 t6[. . . .] iKbpi.ov. P. Lond. II, 350, p. 192, 212 a.d. Offer to lease 150 arourae of shore land, listed in the production category of "dry" land, at a rent of 2 artabae of wheat per aroura. If any part becomes "unflooded" or "waterlogged" the rent will be reduced for the lessees. 4 1 For the reading of 11. 12-14, U.v Si ti Sftooxos f) Ka.8iBa.Tos yivrjTai irapa[5e)x0'j«'a£ Itai, see Preisigke, Berichtigungsliste der griechischen Papyrusurkunden aus Aegypten, Strassburg, 1913, p. 58. 2 Cf. Rostowzew, Bom. Kolonat, p. 188. 3 The resolution Ka0vd(6.Tov) is to be made instead of the editor's KaBiSipovt). 4 P. Lond. II, 350, p. 193, 11. 11-12. 184 W. L. Westebmann C.P.R. 239, 212 a.d. Offer to lease five arourae of shore land belonging to the state, rated in the production category of "dry" land, at two artdbae of wheat per aroura. " If any part of it becomes 'unflooded' or 'waterlogged,' a reduction will be made for us from the rent." 1 These five leases are alike in that they all deal with state land which is rented under the production category of "dry" land. Upon that basis the lessee offers to pay the state a fixed amount. In most of these cases the plots consist of high-lying shore land (ai7iaX6j), which will be covered by the inundation only in an abnormal year. In that case the land would be fiePpeypAvrj. A part of it might remain under water and thus be "unproductive," useless even for hay planting. Against this contingency the lease- holder protects himself by the provision that the naOvdaros portion will be reckoned out of his rent. Also in case this land, ordinarily "dry," should be partially covered by the flood, some part might be rated as afipoxos at the inspection (eiricTKeipis). The inspectors would be compelled in that case to report it in as ajipoxos. The rent, under the government's system of forcing production to the limit of the land that is irrigable, would then be estimated, necessarily, upon the regular high basis of the "unflooded" land. The tenant, who is in these cases the direct lessee, at the time of the paying his rent into the state's granary, pays actually the amount fixed in the lease, showing his lease. The state then takes as a loss on its books the difference between the required high charge on the appoxos and the actual rent paid by the lessee. The lessee has protected himself against the laborious irrigation of the afipoxos. The government has preserved intact its principle of the high rating and enforced production upon this category of land. the kwr\vr\i\pivr\ yfj in leases of public land The eirrivT\iip.£vri yfj has already been defined as that portion of the dfipoxos which has actually been made highly productive by arti- ficial irrigation, whether by using the chain-bucket system or by ditching into the unflooded part. The government's attitude has been explained as an attempt to enforce production to the limit by high taxation upon the "unflooded" category. Proof of the 1 Line 1 1 of this document, wapaSexOri were to be used for hay-cropping. 2 P. Giss. 4, Introduction, p. 24. 3 Archiv fur Papyrusforschung, V, 299. 4 In P. Giss. 4, 5, in the three documents under 6, in P. Brem. 34, and P. Ryl. 96. The corresponding part of P. Lips. inv. 266 is lost. The Clause does not occur in P. Giss. 7, but evidently did occur in another similar lease, of which the ends of the lines appear. Cf. the photographic reproduction of P. Giss. 7 in Klio, VIII, 404-5, where hrqvTXfaiikvris appears in the remnants of another lease, opposite line 16. 5 P. Giss. 4, 6, cols. I and II, P. Brem. 34, P. Ryl. 96: irapa«ex°M*>"7S &0p6xov ko.1 Jliuatias hirovT\T)nkvris Kara t& Was. In P. Giss. 6, col. Ill, Kara rb Wot is omitted. In P. Giss. 5 the clause reads: irapaStxoiikvris ifipbxov xai tirtThrmkuns koto rd Wos, which would lead to the conclusion that the customary payment of half rent upon the "arti- ficially irrigated" was applied here also. 6 Wilcken, Papyruskunde, Grundzuge, I, 1, p. 181. " Unintjndated Lands," Ptolemaic and Roman Egypt 187 book-keeping. This is the customary occurrence, /caret to Wos, as we have seen in previous discussion of the leases. The entirely new feature presented by these leases is that of the payment of half the rent upon the "artificially irrigated" land, the ewqvThrin&ri 777. It must have been a customary payment, /cara t6 Wos. It is to be recalled that the high tax or rent imposed upon the "unflooded land" by the government had the purpose of forcing production upon the land and that this was burdensome in labor upon the tenant. In this form of lease the government agrees that upon that acreage which the tenant does irrigate, 17 lirr\vT\rii).&vi} yrj, he receives a deduction of one-half the rent stipulated by the gov- ernment officials, after the inspection, for his "unflooded land" (a/Spoxos yrj). Just as the rent upon the afipoxos which he did not irrigate was checked back to him in toto by the tax collectors, so one- half of the rent upon the land which he had artificially irrigated was checked back to him at the threshing floor. By this method he was remunerated for the extra labor which he had expended in irrigating the unflooded land. Other than these applications to lease government land upon the especial terms offered by Hadrian's decree of 117 a.d. I have found only one clear example of this type of lease, which must have been quite common before the period of Hadrian and after. It is an offer to lease state lands from the municipal account of Her- mopolis, 1 of the period of Gallienus. The prospective lessee says: "If the land becomes unflooded (af3poxos) from the following year, which I pray may not occur, I will do the necessary artificial irrigation, and I will pay of the above-mentioned rent the half." 2 This was the type of lease which would be at the same time most fair to the tenant, and most advantageous to the government, as a method of forcing production. I believe, therefore, despite the fact that only one instance of this kind has been preserved, that it was one of the standard forms of lease existing in the Roman period. Conclusion. — In the legal theory the landowner is responsible for the land tax, the direct lessee for the rent upon state lands. 1 Stud. Pal., V, 119, col. VII. 2 Ibid., 11. 21-23: Hu> Sk 6 /aij ykvoiro &&poxos -ye'ijTai dird tov l£i)s Itoiw iravayices ewavT\ii6p