But it must be remembered that, in the absence of records, any solution of this difficult problem at present rests on mere speculation and guesswork, and the opinions expressed here must be accepted as mere conjectures unsupported by direct contemporary evidence, and based only upon reasonable probability. We propose to attempt to deal with this difficult subject in the next chapter. CHAPTER IX. _The Succession to the Caithness Earldom._ After the death of Earl John in 1231, we come to a most perplexing time, and it is almost impossible to discover a way out of the maze of genealogical difficulties in which we find ourselves involved. Not only is there no chronicle of the period, but there are hardly any records at all to help us. The pedigree of the descendants of Earl Harold Maddadson, and particularly of his daughters, who are named in the _Orkneyinga Saga_, ceases;[1] and that of Earl John's family and of Harald Ungi and his sisters downwards stops also, save in the case of Ragnhild, the youngest of them, whose son Snaekoll Gunni's son is mentioned as claimant in 1231 from Earl John of certain lands in Orkney and in Caithness as well. Attempts to clear up the mystery have been made,[2] but none of them have resulted in any certain or trustworthy conclusions. Nor can anyone now expect to fare much better; for not only are authentic pedigrees of the Caithness earls and the materials for framing them undiscovered or non-existent, but yet another pedigree, namely that of the Angus line, which provided, from its male members, successors to the title and to a moiety of the Caithness earldom, is very obscure. This chapter, therefore, is largely conjectural, and must be accepted as such. It deserves, and will doubtless receive, severe criticism.