Curse? I'm not aware of any curse, where is a curse coming from? Or at least, I mean, the provision of the lex Julia agraria usually isn't called a curse, in that "curse" in English I guess isn't seen as really meaning the same thing.

I thought this would be sort of a simple question, but when I started going through the sources I found the tradition on the oath written into the lex agraria is pretty complicated and problematic.

The lex Julia agraria (which distributed Campanian land but was emphatically not a "land reform," in that it did not affect land ownership generally) is known from several sources to have contained a clause providing for an oath to uphold it. We do not have the text of the oath, because no copy of the lex Julia survives, but Cicero paraphrases it at Att., 2.18:

habet etiam Campana lex exsecrationem in contione candidatorum, si mentionem fecerint quo aliter ager possideatur atque ut ex legibus Iuliis

The Campanian law contains an oath to be sworn in a contio of candidates for magistracy, if they should make a suggestion that any field should be distributed other than what is provided for by the Julian laws.

My sort of clunky translation is clunky to emphasize several points of the Latin here. First of all, it's likely that the conditional clause that Cicero puts here is a direct quotation, or at least a very close paraphrase, of the text of the law. Roman law very frequently prescribed legal conditions by using conditional clauses. Penalties therefore are usually described in civil law by formulaic conditional clauses in which a hypothetical person is imagined as doing something and then suffering the set penalty, rather than just saying "the penalty for X is Y." Minatory clauses work the same way: rather than saying that something is prohibited, Roman law will often construct a conditional that effectively establishes that the action in question is not legal and will not be considered as such, without actually saying in an indicative clause that you can't do that. Cicero's conditional doesn't fit all that well in the syntax. It doesn't break it, but we'd expect some kind of indirect statement after the first clause, instead of the protasis of a condition acting parallel to the first bit. That suggests that Cicero's quoting or paraphrasing what the law actually says. Note also that there's no apodosis: the punishment is not given. The type of oath provided in the law is given by Cicero, it's an exsecratio. This, I assume, is where the "curse" is coming in, in that somebody is misunderstanding what exsacratio means. An exsacratio was, strictly speaking, an oath that prescribed some sort of divine punishment if broken. I'd have to go track down some other examples, but I'm fairly sure that, as here, the wording of an exsacratio did not generally include the actual punishment invoked, and that instead the oath acted as a conditional with no clear apodosis. In literary Latin the word was sometimes taken to mean something close to English "curse," in that it can be used to describe sort of generally a divine action called against someone or essentially a result clause, as when Cicero describes Thyestes' curse:

ut tu "naufragio expulsus uspiam/saxis fixus asperis, evisceratus/latere penderes"

that you [Cn. Piso, being placed in Atreus' shoes rhetorically] "tossed up from a shipwreck somewhere, stuck fast to sharp rocks, hang cut open at the side."

The use of the ut clause is typical in this sort of curse. But exsecratio in its strict, legal sense, is exactly the sort of thing described by Cicero in his letter to Atticus. First of all, it imagines a precise condition to be violated, without which divine action does not take effect. Second of all, divine action is rarely, if ever, to my knowledge really described--in fact, in most cases of exsacratio known in legal contexts the oath incurs a legal penalty, not a divine one, as if the law was prohibitive in nature. These two facts separate it from, say, the curses that we find in curse tablets or in say the formula of devotio as related by Livy, in which the god is called on to turn his wrath, often in a very specific way, on the target, frequently with some sort of imperative or result clause. In the case of the lex agraria it's not clear whether the punishment for violating the condition described was actually spelled out. In this case there seems to have been some provision for legal action of some kind, since all our other sources mention that there were harsh penalties attached to it. I'll get back to that later, it's not terribly relevant right now. Lastly, and importantly, exsacratio is not a curse in that it is issued by a speaker against a target, general or specific, but rather is an oath that is sworn, with the punishment for violating it assumed by the swearer--and only by the swearer--once it is sworn. So in the next sentence at Att., 2.18 Cicero specifically says that Laterensis, who was then candidate for the tribunate, resigned his candidacy so that he would not have to swear the oath.

That leads us to another problem, namely the target of the clause. Cicero says that the oath was to be sworn in contione candidatorum, "in a contio [i.e. a public address to the people] of the candidates for office." That has a couple of implications. First of all, Cicero restricts the oath only to candidates. Senators, sitting magistrates, etc. do not have to swear it. The implication seems to be that the magistrates for the next year would not violate the conditions of the law. I'll get to the other sources in a bit, but it's not clear what exactly the clause provided was not to be violated. Cicero's apparent quotation quite clearly says that the law provided that the magistrates for the next year were not to distribute ager publicus (which is what's meant by simply "ager") other than the ager Campanus which the lex Julia portioned out. Whether the idea is that Caesar feared that the magistrates for the next year might reassign the settlers before their plots were set up or that he was actually trying to limit the threat of populist land distribution (thereby making the law more palatable to the opposition? The resignation of a candidate for the tribunate, precisely the sort of person who would be expected to be targeted if this is the purpose of the clause, is suggestive) I'd have to think about, and check up on the scholarly analyses of this clause. Again, this will matter when I get to the other sources, but relevant right now is the fact that Cicero describes the oath as applying essentially only to the magistrates of the next year and only in a specific way, not as a general protection of the law's validity. Secondly, the mention of a contio seems unusual. We know the text of a Lex Latina Tabulae Bantinae which includes an oath to be sworn by Quei senator est eritve inve senatu sententi]am deixerit post hance legem rogatam ("anyone who is a senator or expressed his opinion in the senate after the passage of this law," which is to say future senators and speakers in future debates on the law) to uphold it (more on this later), and it stipulates that various magistrates are supposed to swear in front of the Temple of Castor, facing the forum, and that within five days the magistrates are to swear to a quaestor that they'll uphold the law and all that. The text is fragmentary, but it seems that the idea is that the magistrates are supposed to swear before the people and then before the quaestor, or something to that effect--a similar oath was sworn by magistrates to respect tribunician sacrosanctity. This is another case where I'd have to check to see whether such things are described as contiones normally, but that seems somewhat odd, given that contiones are essentially speeches to the people--in any case Cicero seems to be describing an oath sworn to the people. Given these two facts it seems a reasonable reading to interpret what Cicero is saying as some sort of provision to make sure that the magistrates, ideologically beholden to the electorate that granted them their office, carried out the provisions of the bill properly as an instruction of the citizen assembly.