This note is based on two works:
Aside from a few statutory dispositions, the primary source for the law governing family and inheritance among reichsständische(immediate) families of the Holy Roman Empire consisted of either customs, or else family pacts, wills, house laws, etc. These families essentially enjoyed considerable autonomy, often broadened, never restricted. In the absence of specific sources, the applicable law was primarily the German common law, and secondarily Roman law ("received law") as common law of the Empire. As immediate members of the Empire, these families were subject to local law only to the extent that they willingly submitted themselves.
The local estates had no customary role in the formulation of house laws, nor was their publication a condition for their validity. Confirmation by the Emperor was necessary only to the extent that the dispositions of the house laws infringed on the reserved rights of the emperor or on the law governing imperial fiefs (Reichslehnen); or when rights of family members were restricted or abrogated where common law of the Empire recognized them as unalienable without consent. For example, the rights of the unborn, the nascituri, as well as the rights of the underage, came under the jurisdiction of the Imperial court, the Reichshofrat; also, house laws depriving daughters of their reserved portion or formulating ebenbürtigkeit requirements were submitted for imperial approval. In fact, there is no known case of introduction of male primogeniture without imperial approval. Even then, imperial confirmations were not routinely published.
After the end of the old Reich in 1806, the next constitutional framework was provided by the deutsches Bund. But the laws of the Bund were silent on the subject of house laws. As a result, the limitations on families' autonomy that existed under the old Reich disappeared. Furthermore, even the introduction of constitutions in some states did not necessarily make consent by the estates or parliament a requirement (the constitution of Hannover of 1840 explicitly says that the house laws do not require the consent of the estates). In some cases (Württemberg, Saxony, Saxe-Coburg-Gotha) consent of the estates was sought for some articles of the house laws. However, according to Zoepfl, that does not mean that it became possible in law for the head of the house to deprive agnates of their rights without their consent, even with the consent of the estates. (He notes the difference with English law, where the king jointly with the houses of parliament can alter the succession as he wishes; he does concede that, after 1806, there was no court where agnates could seek protection of their rights).
For Rehm, a house of the Hochadel had a legal persona, similar to a corporation. The ability to make statutes did not belong only to its head, but to all members.
There was a period, he concedes, when house laws were passed without consent of the agnates: that is the period of absolute monarchy following the end of the old Reich in 1806. This was founded on two views: 1) one is that house laws were really state regulations, laws that applied to the ruling house rather than house laws. Where the state exercised its authority, consent of the affected parties was not required (this is, of course, in the context of absolutism). An example is the house law of 4 Oct 1817 in Baden, issued in acknowledged violation of the house law of 10 Sep 1806, "kraft der Uns zustehenden Souveraineté" (by virtue of our rightful sovereignty). The Hanover house laws of 1833 and 1836 are in the same spirit. 2) the other is that the ruler is sovereign in his house just as he is in his state. Here, the house law is house law, but the head of the house has absolute power. This is the spirit of the Hohenzollern houselaw of 24 Jan 1821, issued "kraft der Uns als Oberhaupt Unserer Fürstlichen Familie zustehenden Gewalt" (by virtue of the power which is ours as head of our princely family). Likewise the Bavarian houselaws of 1808 and 1816. The family statutes of Baden of 1806 combine both views (giving succession rights to the issue of his second marriage).
Either view limited the ability to bypass agnates' consent to the degree that the agnates were subjects of the head of house. That is why the Hohenzollern laws of 1821, and the Hannover laws of 1836, required the explicit consent of the Sigmaringen line and the Braunschweig line respectively.
Either view, as basis of the absolute power to make house laws, went hand in hand with absolutism, and therefore disappeared as constitutions were introduced. This prompted a return to the historical view that the consent of agnates was required; the only change was in form, whereas earlier a will (where only one's issue was concerned) or a family pact was used, now a houselaw was issued but with consent of the agnates.
The speed of the change is indicated by a couple examples: the Bavarian houselaw of 1816 was issued without agnatic consent; the constitution promulgated in 1818, and the houselaw of 1819 "unter Zustimmung der Agnaten erlassen". Likewise, the three Reuss princes issued rules in 1844 on their own; the Reuss junior line submitted to a constitution in 1849, and the house laws of 1855 explicitly cited the agreement of all agnates. Two house laws actually included the requirement of agnatic consent to any future modification: one was the Hanover house law of 1836 (although itself promulgated by virtue of the sovereign's power only), the other was the Oldenburg statute of 1 Sep 1872, itself approved by the Tsar and all princes of the grand-ducal house with succession rights, also specified (art. 71): "Änderungen dieses Hausgesetzes können von dem Großherzoge nur mit Zustimmung sämtlicher dem Familienrat angehörigen Mitglieder des großherzoglichen Hauses vorgenommen werden".
There was, however, a later case where house laws were changed without agnatic consent. As a result of the dispute over the Lippe succession in 1895-97, doubts about the Ebenbürtigkeit of the Lippe-Biesterfeld line had been raised. The third son of the duke of Saxe-Meiningen had married someone from that line. A law was passed on 9 March 1896 declaring her ebenbürtig for the purposes of Saxe-Meiningen. The consent of the duke's sons had been obtained, but not cited in the law itself. The reason was that, in truth, the consent of other Saxon lines would have been required. The government's view, however, was that "the duchy was a sovereign state, in contrast with the days of the Holy Roman Empire, perfectly capable of setting laws to the effect that the issue of this marriage is dynast in Saxe-Meiningen. There exists no legal mean in Germany to dispute such laws, promulgated through a constitutional process" In Rhem's view, this is compatible with his thesis: to make the law perfectly valid would have required consent of all agnates, but as a matter of positive law there is no way to prevent its application.