OP 6s £4 Pete er erere sete op cere te bdetets ere arty. Pew Cus f ay. tt hi 4} i) rangi ete , ices Hitinte ~ < =ié ~ > A m 1 wer PROCEEDINGS ROYAL IRISH ACADEMY VOLUME XXXIV DUBLIN: HODGES, FIGGIS, & CO., LTD LONDON: WILLIAMS & NORGATE 1917-1919 ‘ THE Acw VCR ACAGS@ a + OF GolEN PROCEEDINGS OF THE ROYAL IRISH ACADEMY VOLUME XXXIV SECTION A.—MATHEMATICAL, ASTRONOMICAL, AND PHYSICAL SCIENCE. DUBIN SOG ES. BIGGS & CO. Trp: LONDON: WILLIAMS & NORGATE 1917-1919 Tue ACADEMY desires it to be wnderstood that they are not answerable for any opinion, representation of facts, or train of reasoning that may appear in any of the following Papers. The Authors of the several Essays are alone responsible for their contents. Dupin: Printep at tHe Untvenstty Press ny Ponsonby AND Gieis. CONTENTS SKCTION AWMATHEMATICAL, ASTRONOMICAL, AND PHYSICAT, SCIENCE. PAGB Hirowcoox (Frank L.), Px.D., :— On the Simultaneous Formulation of two Linear Vector Functions, . 1 Leatuem (J. G.), M.A., D.Sc. :— On Two-dimensional Fluid Motion, with Free Stream-Lines, past an obstacle of Curved Outline, ; : F j : : a ll McCretuanp (J. A.), D.So., F.R.S., and Rev. C. J. Power, 8.J., M.Sc. :— Electrification by Friction, . 5 : : . , F . 40 McCuettanp (J. A.), D.So., F.R.S., and P. J. Nozan, M.So. :— The Ions produced by bubbling Air through Alcohol, : é 5, lal! Notan (P. J.), M.Sc. :— See under MoCuetuanp (J. A.). Prummer (H. C.), M.A. :— On the Symmetrical Optical Instrument, : ° : : 5 ry Power (ev. C. J.), 5.J., M.So. :— See under MoCretuanp (J. A.). Roserts (Rev. W. R. W.), D.D., S.F.T.C.D. :— On the Equation of the Tangent at a given point on a Uni-nodal Quartic Curve, 6 5 4 : 5 5 ; : a G8 PROCEEDINGS OF THE ROYAL IRISH ACADEMY PAPERS READ BEFORE THE ACADEMY I, ON THE SIMULTANEOUS FORMULATION OF TWO LINEAR VECTOR FUNCTIONS. By FRANK L. HITCHCOCK, Pu.D., Massachusetts Institute of Technology. {Read January 22. Published May 25, 1917.] CONTENTS. PAGE PAGE 1. Introduction, . : 9 : 1 7. Typical form of » and @ when »~-1@ 2. Possibility of the reduction toform (1), 3 has an infinite number of axes, 3. Typical form of » and @ when »~ 16 8. Special case of self-conjugation, has a double axis, < 3 9. Typical form of @ and @ when no 4. Geometrical application, . 0 : 4 function of the pencil has an 5. Typical form of » and @ when o~ 10 | inverse, : ~ # » 7 has a triple axis, . : 5 10. Application to quadratic vector func- 6. Restriction to self-conjugate func- tions, : ¢ - : 5 8 tions, . . Ci 3 6 F 5 | 1. Introduction.—In his valuable appendix to Hamilton’s “ Elements of Quaternions” the late Prof. C. J. Joly exhibited, in a very elegant manner, some of the properties of the most general pair of linear vector functions. He pointed out that two such functions @p and Op may, in the general case, be simultaneously expressed in the form pp = \Sap + wS3p + vSyp; Op = wASap + buSPp + evSyp, (1) where the three vectors a, 9, y are the axes of @’¢’~', while A, uw, v are the axes, and a, b, ¢ are the latent roots of p@~'. These function-pairs are important in many ways, and appear frequently in Joly’s writings. In the above formulation we note that he has expressed R,1,A, PROG., VOL, XXXIV., SECT. A. (1) 2 Proceedings of the Royal Irish Academy. @ and @ in terms of those vectors whose directious are similarly altered by the two functions. Sometimes, as in his paper* on “Scalar Invariants of Two Linear Vector Functions,” ke finds it more convenient to express } in terms of its own axes, and then to formulate @ m terms of ¢. Two of his invariants will be useful in the present investigation. On the purely analytical side, the problem of the present paper is analogous to that presented by the classification of pairs of bilinear forms, or of pairs of collineations.+ To the worker in quaternions, however, a purely scalar treatment is unsatisfactory, first because the physical and geometrical significance of the results is pretty thoroughly concealed by the method of presentation, also because relations of singular and non-singular forms to each other and to the invariants of the system cannot be, or at least has not been, clearly brought out by ordinary algebra. A singular bilinear form (or a singular collineation) corresponds to a linear vector function one of whose latent roots is zero. It will appear below that either or both of the given forms may be singular without altering the typical properties of the system : in fact, it is only when the occurrence of simultaneous vanishing roots is accompanied by the vanishing of two of Joly’s invariants that the system falls into a more special type. To mention a problem of another sort in which Joly’s formulas for ¢ and 6 prove to be of great assistance: the general quadratic vector function of p can be written Vop0p + pSéep, (2) where 6 is a constant vector—a fact bearing on the theory of certain functions defined by differential equations. In factorizing a quadratic vector in this way, simple methods for the simultaneous formulation of ¢ and @ are necessary. Again, to take a problem from geometry, if a curve of the fourth order in space be given by the intersection of two quadric surfaces, the equations of the curve may conveniently be written Spop + Sap + a = 0, i Sp6p + SBp+b=0, | where ¢ and @ are self-conjugate linear vector functions, a and PB are constant vectors, and @ and 4 are constant scalars. Many of the properties of the curve appear most clearly when ¢ and 6 are expressed by formulas analogous to (1). (3) *Trans R.I. A. 30 (July, 1895), p. 709. + For a short account in English of this aspect of the matter, see Bocher and Duyal’s “ Introduction to Higher Algebra,” chap. xxi. } For a sketch of this theory see ‘“‘ A Classification of Quadratic Vector Functions.” Proc. Nat. Acad. of Sciences (Washington), vol. i, No. 3 (March, 1915), p- 177. Hivencock— On the Formulation of two Linear Veetor Functions. 3 2. Possibility of the reduction to form (1).— In all these problems we are concerned with the properties of a pencil of functions +40. That is to say, in studying the curve (3), or in forming the vector product V¢p0p, or in calculating Joly’s invariants, we may replace a given » or @ by any other functions of the pencil. The general problem is, then, to determine what types of function-pairs may occur. In writing ¢ and @ in the form (1), Joly assumes the possibility of the reduction. To see on what the possibility depends we note, first, that Vy, Vya, and Va are the axes of $-'@) A necessary condition for reduction to the form (1) is, therefore, the existence of three distinct axes for @7'0. Second, if @ has a vanishing root, so that $-'0 does not exist, we may suppose @ replaced by a different function of the pencil. But it may happen that no function of the pencil possesses an inverse. ‘The reduction (1) is impossible in this case also. We have therefore to examine two cases :— I. The function ¢'0 exists, but has a double or a triple axis. II. The function ¢7@ cannot exist, ie. no function of the pencil @ + 40 possesses an inverse. 3. Typical form of » and @ when ~-'0 has a double axis. Suppose » 4 to have a double root. As I have shown in a former paper,* we may then write, as the most general form of ~'6, @ '0 = gp + cBSPBip + 4B SaBp, (4) where g is the double root of the cubic in $~'0, and may vanish. The double axis is 3. It is assumed that ¢, ¢,, and Sa/3(3, are all different from zero. ‘lo express ¢ and @ in a simple manner we have now merely to operate by @ on the three diplanar vectors a, B, 8,. Let the results be denoted by 2X, w, v, respectively ; and these three vectors are also diplanar, because, by hypothesis, » has an inverse. Expanding ¢ in terms of A, pn, v we have op - Sab. = MPBip + wSB,ap + vSaBp, (5) and by operating with ¢ on both sides of (4) we have Op = pp + cwSB3,p + ewSapp (51) These expressions actually differ little from (1). To bring out the analogy we first note that (5,) is equivalent to the three equations Oa = gr + cuSaf33,, Je) = IM Op, =(g + ¢,SaBp,)v, * Proc. Royal Soc. Edinburgh, vol. xxxv, Part II (No. 17, June, 1915), p. 172. (1*) 4 Proceedings of the Royal Irish Academy. so that if we agree to write cSaBBi=c. and y + ¢SaBP, =9:, we may rewrite @ in terms of its effect on a, 3, and ,, thus— Op . Saf3(3, = (9X + cou) SBPip + JuSB.ap + gvSaBp, (52) where the analogy with (1) is more obvious, g and g, being the two roots of the cubic in ¢'6. To complete the formulation, put a,Saf33, = VBp., 2SaBp, = VBya, 7:Sa3pi = Vaf3, whence (5) and (5,) become pp = ASa:p + wSB.p + vSy.p, Op = (gr + Cou) Sasp + guSB:p + gwSyzp 6) It is clear that the form of ¢ is as in (1), while @ differs only in its effect on a. 4. Geometrical application.—If equations of a curve are given as in (3), we write down the conditions for self-conjugation from (5) and (5,)— VrAVBB, + VuVBia+ VoVoB=0, VeuVBB, + VawVaP = 0, which (by Hamilton’s Elements, Art. 294) are equivalent to the six scalar equations SBiu = SBv, Sav = SBA, SBA = Sap, c, SBv = 0, ¢,Sav = cSB.n, cSBu = 0. To solve, regarding a, 8, and PB, as known vectors, we note first that c and c¢, are by hypothesis different from zero. The six equations thus reduce at once to SBu=0, Spu=0; Sav=0, SBv=0; SBA =0, and SBA = Sap Hence yu, being perpendicular to both 8 and A, has the form pVP, where pis some constant scalar. Similarly » has the form gVaj3, and X has the form 7 V3, + pVB,a, where g and 7 are two more constant scalars. Collecting results, (5) and (5,) become dp. Sa88, = (r VBB, + pV.) SB/3.p + p VBB,. SB,ap + qVaf3 . Sa3p, (7) 9p = 99p + PVBPSBB.p + oq Vaj3Safsp. The self-conjugate character of these two linear vector functions is evident from their form. And since, in the geometrical problem, any multiple of one of the quadratic expressions (3) may be added to the other without altering the curve of intersection of the two quadric surfaces, we may take g = 0 in (7). If we now operate on (7) by Sp and introduce a set of oblique coordinates defined by the equations #SaBp.=SBB.ip, ySaj3pi = Sap, — Saj3j3. = Sajsp, Hirencocx—On the Formulation of two Linear Vector Functions. 5 we shall find the scalar products Sp¢p and Sp@p to be proportional to the quadratic expressions ra + Qnay + gz and cepa + agz*; If we now combine the two equations so as to eliminate 2* from the former, we may obtain the terms of the second degree as ac’ + 2pey and pa + qe, where a, p,, and g, are constants. The form of the result shows that the curve has a double point at infinity. 5. Lypical form of » and 0 when $0 has a triple axis.—Suppose p70 to have a triple root. The most general form of ¢6 is* @ Op = gp + ¢BSByp + crySyap, (8) where g is the triple root, and may vanish. It is assumed that ¢,¢,, and Say are all different from zero. As before, we regard a, (3, and y as known, operate on them by @, and call the results A, uy, and vy. Whence pp Say = ASByp + wSyap + vSa[3p, (9) and by operating with @ on both sides of (8), Op = 9hp + cuSByp + civSyap. (9,) This is, perhaps, the most convenient form of #, but we may, if we wish, bring out the analogy with (1) and (6) by writing C2 = cSapy, Cs; = oSapy, a;Sapy = VBy, BSafsy = V By, and yaSa[sy = Vaf3, when we shall obtain bp = ASasp + wSPsp + vSysp, \ Op =(gA + Com) Sasp + (gu + ¢3v) SBsp 3 guSysp- (10) It is clear that @ differs from (1) in its effect on a and on 3. The only direction similarly altered by ¢ and @ is y, the triple axis of $70. 6. Restriction to self-conjugate functions.—lHt, as in Art. 4, we require that ¢ and @ shall be self-conjugate, we have by (9) and (9,) VAVBy + VuVyat+ VoVap=0; VeuV Sy + VewVya = 0, which, ¢ and ¢, not being zero, are equivalent to SBv =0, Syv=0, Syw=0, Sav= Syd, SBA=Sapn, and cSBu=cSav. Solving for X, u, and v we have n=? VBy +q Vya + ep Vaf3, M=q VBy + Cp Vya, v = cp VBy, * Loc, cit., p. 174, equation (24). 6 Proceedings of the Royal Irish Academy. where p, 9, 7 are three arbitrary constants. If we substitute these values of A, u, v, in (9) and (9,), the self-conjugate character of @ and 0 is evident. 7. Typical form of 6 and @ when $*6 has an infinite number of axes.—It may also happen that ¢“@, whether possessing a double axis or not, has an infinite number of axes. The most general form of ¢°6 is then* @ ‘Op = gp + ySep, (11) where g is a double root of the cubic in #70 (not necessarily corresponding to a double axis), and y and « are constant vectors. Let a and 3 be two vectors which. with y, form a diplanar system. Call ga, ¢B, and gy, as before, A, nu, and v, respectively. Expanding gp we have op - SaBy = ASByp + wSyap + vSaj3p, (12) and by acting on (11) by ¢, Op = gdp + vSep. (12) It thus appears that ¢ and @ alter in the same way the direction of y, and also that of any vector perpendicular to « If y is itself perpendicular to «, the function ¢'@ becomes a “shear,” and y becomes a double axis. If we require » and @ to be self-conjugate, we may take @ any self- conjugate linear vector function whatever, and must have e parallel to v. Operating by Sp, and setting g = 0 in (12,), we see that SpOp is a perfect square, or may be made so by a proper choice of quadrics from the’pencil. We may, if we wish, regard various special cases of (12) and (12,) as being merely limiting forms of the function-pairs (1), (6), and (10). For example, if either c or c, vanishes, (8) falls into the form (11), and @ may be written as in (12). 2 8. A special case of self-conjugation—The foregoing function-pairs have been built up by assigning particular forms to g¢'@. Various interesting special cases may be obtained by assigning some special form to @. As an illustration, important in the theory of curves of constant torsion, let @ be required to have a double axis and to be self-conjugate. It is clear that » cannot be real, since a real, self-conjugate linear vector function has three distinct, mutually perpendicular axes. In fact, if we write ¢ in the form of the right member of (4), which is the most general form of a linear vector function having two coincident axes, viz. :— $P = Jp + cBSB/3:p + 43,Saj3p, the condition for self-conjugation is eV BV BB: + 4.V 2B, Vas = 0, * Loc. cit., p. 175, equation (28). Hircacock— On the Formulation of two Linear Vector Functions. 7 equivalent to the three scalar equations GSBB = 0) eSabr— cSBE, = 0 4 cb? = 0. If ¢ and ¢, are different from zero, these reduce to SBB, = 0, Saf. =0, PB = 0, whence /3 must be a vector whose square is zero, i.e. a minimal vector. An example of such a vector is i+ j,/-— 1. We may satisfy the conditions by taking 8, =, and a=7, with ¢ and ¢ any constants. We shall then have pe = Jp + aG oF iu) S(- i] 3 5 fy =al)) p + ekSkp. Here 7, j,k are any three unit vectors forming a rectangular system, ¢ is self-conjugate, and has the two axes k and i+j7,/-— 1, the latter being a double axis. When ¢p has been developed in this or any other manner, a second function @p may be formulated, following Joly, by expanding @ in terms of three vector constituents of ¢. Thereby @ is determined by means of nine constants, and it is in terms of these constants that Joly expresses a large number of invariants (cf. note 1). 9. Typical form of » and 0 when no function of the pencil has an inverse.— The function-pairs (1), (6), (10), and (12) include all possible cases except when no function of the pencil possesses an inverse. In this latter case, the cubic in » + #0 must have a vanishing root for all values of ¢. Now the constant term in this cubic, as was pointed out by Joly, may be written ms + tly + Cl, + tM, where ms; and ms are the third invariants of » and @ respectively, and l/ and /; are Joly’s new invariants given by 1, = S&pa0 Oy + pBOyOa + gpybah) 1, _ S(BapBoy + OBdyga + Oypagp) Sapy : Saf3y The invariants m;’ and i; vanish by hypothesis. If, and only if, /;’ and J, also vanish, we shall have no function of the pencil possessing an inverse. These quantities are invariant in the sense that their values are inde- pendent of our choice of the three vectors a, 3, y, provided they are diplanar. Let a be chosen to be the direction annulled by ¢, that is ga=0. We shall have two sub-cases, according as @a is, or is not, zero. If Oa = 0, it is evident that a is zero for every function of the pencil. If Ga is not zero, let (8 be chosen to be the direction annulled by @, ie. 68 = 0. Let y be any vector not coplanar with a and 3. Joly’s invariants now become /,= SpPey4a Ae ie SOappoy i ; Say F 8 Proceedings of the Royal Irish Academy. Tf these are both zero, we shall therefore have simultaneously S70 Ve0ag8=0 and Sy¢’ VOagB = 0. But y was an arbitrary vector, hence these equations imply VGa¢g5=0 and 9’ V6agB = 0. that is, the vector V@ag3 is a zero for both of the conjugates ¢’ and @, This direction will then be a zero for the conjugate of any function of the pencil. The two sub-cases may accordingly be characterized as follows :— 1. ¢ and @ have a common zero. This implies that the conjugates ¢’p and @’p lie in a common fixed plane for all values of p. 2. @’ and 6 have a common zero. This implies that op and 6p lie in a common fixed plane for all values of o. The second case is easy to formulate. Since the common fixed plane is known when ¢ and @ are known, we may choose i and 7 any two perpen- dicular unit vectors in that plane, and expand ¢ and @ in terms of them, when the two functions necessarily appear in the form oe = iSaep + jSanp, Op = iSBAp + jSBup, (13) where «, n, A, and are four constant vectors; for, by hypothesis, ga = 0 and 68=0. To find « we have — Si¢p = — Spi = SpVae, whence, p being any vector, we must have Vae = - ¢’i; therefore « may be any vector at right angles to »’i, distinct in direction from a, the tensors of a and of « being selected to satisfy this equation. In a similar manner we may find values for 7, A, and u. In the other sub-case ¢’p and 6’p lie in a common fixed plane. Hence they must be, in form, the conjugates of the right members of (13) or some equivalent. That is, we may write for this case op = VaeSip + VanSjp, 4p = VBASip + V BySjp, (14) where ¢, n, A, and w are to be found by treating the conjugates as ¢ and @ were treated in (13), i and j are taken in the plane of the conjugates, & is the common zero of ¢ and @, and a and # are the directions annulled by the conjugates, 9’ and 6’, respectively. The forms (13) and (14), from their method of formation, cover all function-pairs such that no function of the pencil has an inverse. 10. Application to quadratic vector functions.—In conclusion I shall prove Hirexcock—On the Formulation of two Linear Veetor Functions 9 a theorem, which, as I have elsewhere shown,* is fundamental in the theory of quadratic vector functions. Theorem: The locus of the IRREDUCIBLE+ vector VppOp cannot be a fixed plane. To prove the theorem, we note that by the preceding investigation all possible function-pairs may be written in one or the other of the six cases, (1), (6), (10), (12), (13), or (14). Forming the product V¢p0p from (1), we have Vop0p = (c-b)VuvSBpSyp + (b- a) VAuwSapSBp + (a - ¢)VurSypSap, (15) where A, », and v were by hypothesis diplanar, since @ has an inverse. Hence Vuv, VvA, and VA are also diplanar. Furthermore, a, b, and ¢ are unequal, since, in this case, p70 has distinct roots. And a, 3, y are diplanar for the same reason. Hence the locus of (15) cannot be a fixed plane, and the theorem is proved for this case. Forming the product V@p0p from (6), we have Vop9p = (g. — 9) VuwSBpSyp + (9 - 91) VuASypSap + @[ VAwS*ap — VuvSypSap], (16) where similar reasoning holds, viz., A, 4, and v are diplanar because ¢ has an inverse: hence the vectors Vuv, VvA, and VA are diplanar; g —g, 1s not zero, since @'@ has two unequal roots; ¢, was by hypothesis not zero; and, finally, the scalars Sap, SBp, and Syp cannot be multiples one of another, since a, 3, and y are diplanar. Hence the locus of (16) cannot be a fixed plane. Forming the product V@p8p from (10), we have V ppOp = a VuvS*yap - VivA SBypSyap | + of ViwS*pyp = VuevSBypSyap |. (17) Here neither ¢ nor ¢, can be zero under the hypothesis for this case. The rest of the reasoning is as before. Hence the locus of (17) is not a fixed plane. Passing to (12), which is equivalent to removing the restrictive hypotheses on the e's and g’s of the former cases, it is evident that we have V¢pOp divisible by Sep, hence reducible. From (13) we have Vp0p parallel to &, and so divisible by a quadratic scalar, that is reducible. * Proc. Nat. Acad. Sci., Joc. cit. Fora more detailed study of quadratic vectors see ‘© 4 Classification of Quadratic Vectors.” Proc. American Acad. of Arts and Sciences, 52-7 (January, 1917), p. 369. + An irreducible vector is one which cannot be factored into a variable scalar and a vector of lower degree,—a term of the form pé being added to the vector if necessary. As an equivalent definition we may say that a vector Fp is reducible when, and only when, the vector product Vp/'p is divisible by a scalar variable. R.I,A. PROG., VOL. XXXIV., SECT. A, [2] 10 Proceedings of the Royal Irish Academy. Finally, from (14), forming the product V¢pfp we have Vipp0p = Sip V VaeV BX + SipSjp[ V Vae VBu + VVan VBA] + Sip V Ven VBu. (18) Since 7 and j are at right angles, it is only needful to show that coplanarity of the vector coefficients entails the reducibility of the expression ; in fact, to show that whenever we have S.VVaeVPAl VVaeV Bu + V Vay VBA\V Ven V Bp = 0, (19) we also have the right member of (18) divisible by a scalar factor. The scalar product may be expanded as ~ 8 .VaeVBAV BuS .VaeVanV Bu + S.VaeVBAVanS .VBAVayV Bu, (20) by applying the identity S.Vo:p:V psp. Vesps = SpipspsSpupsps — SpipresSpspspe in which we write p,;= Var, p:= VA, &c. Expanding again in a similar manner (20) may be written — SagGSAByu . (— SaeySaf3u) + SaenSaBr . SPA Sanp. (21) Now if SAP = 0, the direction of Mp in (14) is constant, hence Op is divisible by a linear scalar and (18) is reducible. Similarly, if Saen = 0, Op is reducible. Rejecting these factors from (21) we have SacBSapu — SaBrASanf = 0, (22) as the only remaining possibility. But this is the condition that ¢’B shall be parallel to 0’a; for, remembering that (13) is the conjugate of the present case, we have p/3 = tSac3 + jSanj3, Oa = iSBra + jSBua ; (23) the condition that the coefficients of i and j shall be in proportion is equivalent’ to (22). Write, accordingly, 6’a = 79’j3, and take y any vector not coplanar with a and 8. Identically pSapy = aSByp + BSyap + y Sap, whence, operating with 9’ and 6’, and remembering ga = 0B = 0,— &'pSasy = $ PSyap + ¢ ySaf3p, WpSapy = rp BSpyp + W'ySaf3p. By taking conjugates of both sides ppSapy = VyaSq' 3p + VaSo'yp, OpSaBy = 2VByS¢' Bp + VapSo'yp. Forming the vector product we have Vp . S'apy = S¢'pp(zV Vya VBySt'3p +VVya Vapse'yp + 0. This implies a relation between w and @ such that the relevant region in the w plane is also conformally represented upon the half-plane »>0. Thus there are two functional relations, the “geometrical relation” between 2 and Z, and the “field relation” between wand. Figure 3 represents the @ diagram, and it is convenient é =0o9 —@ c a + oo Fic. 3. to specify points of interest in the other diagrams by the values assigned to € at these points, namely ¢ at the prow, —@ and @ at the points of departure of the free stream-lines, and — o and + c for infinite remoteness along these stream-lines. A suitable form for the field relation is w=4V (0); (1) the main problem is to find a geometrical relation which shall satisfy not only the geometrical data but also the special requirements of free stream-lines. When this problem is dealt with by the method of curve- factors no variables are required but those above specified. 3. If the obstacle is symmetrical about a line parallel to the ultimate stream, the distribution of flow is symmetrical about the same line. Such a configuration is typified in figure 4. In this case it is convenient to adopt a to ae ee in iano Fic. 4. different formulation and to regard half the field of flow as constituting the relevant region in the z plane, the boundary being completed by the straight Stveam-line in the line of symmetry. The corresponding region in the w plane is the half-plane on the positive side of the real axis, so that the intermediate variable Z is not required, and w being equal. The value assigned to % or w at the prow is now zero, and the value at the point of departure of the free stream-line is a. Learaem—On Two-Dimensional Fluid Motion. 15 4. It is convenient to put uw + ww =q exp (7a), so that g is the resultant velocity and a its direction-angle, and to noticé that dw/dz=u—-iv. The characteristic of a stream-line that is free is that, along it, the resultant velocity g or | dw/dz | is constant. If the geometrical relation be dz = G62) dé, (2) where (,(Z) isa curve-factor, the constancy of g involves that of | 6,(Z) d&/di |, so that, if w=4V(€-c)*, | 6,(%)/(€-e¢) | is constant. Along a free stream- line which extends to infinity the constant value of g must be V, and therefore the constant value of | 0(&)/(¢ - ¢) | is unity. If, as in the case of symmetry, w = Z, then | 60/2) | is constant along a free stream-line. GENERAL FORMULAE. 5. A field relation and a geometrical relation of the forms of equations (1) and (2) respectively having been assumed, the problem of formulation is reduced to that of finding a suitable general formula for ¢,(Z). The requirements which (,(Z) has to satisfy are : — (i) that it shall give a conformal relation between the relevant regions of the planes of z and Z, which is equivalent to saying that ii must be a curve-factor, not necessarily pure (G, § 1), whose angular range is 7, and (ii) that all along the parts of the boundary corresponding to free stream-lines, namely the parts where @ is real and @? > a’, the modulus of ¢,(2)/(Z -— c) shall be unity. In the case of symmetry there is a single relation, say dz = Gi (w) du, (3) and what is required of G,(w) is that it shall be a curve-factor of zero angular range, whose linear range is the part of the real axis of w in which w is positive, and such that the modulus of ()(w) is constant for w real and greater than a. Whether for ¢,(Z) or G5(Z), the general formula is a product or limit of a product of curve-factors of suitable type and possibly also Schwarzian corner-factors. But there is a considerable range of choice of the type of sub-factor. In the case of a limit of a product it is convenient to represent 6, or @, as the exponential of a definite integral, making use of equalities of the type k=) F(x) 5x b Lim [J {@(«)} = Exp | /(«)log{@(«)} de. k= a 16 Proceedings of the Royal Irish Academy. 6. In passing it is worth while to refer to possible types of sub-factor other than that which it is proposed to adopt. Adverting to the case of symmetry, it is known (D, § 37) that, when « is a real parameter, that branch of {tx + (w — x)? } /w?, in the relevant region of the plane of w, which > 1 for w~>+o is a curve-factor of zero angular range, representing a straight boundary for w real and 0 < w «, and having constant modulus on the curved boundary. Hence : a ae Cy (w) = Exp | log [}ik + (w-x) }/w | f(x) de, (4) 0 where 7 is any function of the real variable x, is a form of G,' which satisfies the special requirements. Save for slight changes of notation and sign- convention, this is the same as w?(.;(w) as detined in D, §40. It is mentioned here because, if Mr. Levy's analysis (paper E) were replaced by an equivalent analysis in terms of curve-factors, it would prove to correspond (in the case of symmetry) to using a form of @,’(w) which differs from the above only as the product of a definite number of factors with definite indices differs from the exponential of a definite integral, which is its limit when the number of factors becomes great without limit and the index of each factor infinitesimal. Similarly, Mr. Levy’s treatment of the asymmetrical case would correspond to the use of powers of sub-factors of the type | (B-a~2e)(Z ~<) | 2(a+ c)}(B - c)? which is a particular case of (, and (,; as defined in F,§8. The index would ultimately be a function of a and (3, while these parameters would take values respectively between -c¢ and a, and between ¢ and a. GAG) =i + (a+ (PB - | +(€+a)Z-p), (5) Some reason for doubting the complete generality of @,/ as given by formula (4) is suggested in Article 18 below. When Prof. Levi-Civita’s analysis (paper A) is replaced by a parallel analysis in terms of curve-factors, it is found to be essentially equivalent to representing (2) as the product of a single factor {G3(Z) (Z = c)}??, in which a = 8 = a, and a number of sub-factors of the type Gat) = Exp [- ic,a-” | — rape (¢ - a2) y"), (6) nm taking integral values. On comparison with D, § 43, it will be recognized Learaem— On T'wo-Dimensional Fluid Motion. 17 that @,, belongs to the type of (;, but is closely akin to 6@.. and has zero order at infinity. : Still another type of sub-factor is the foundation of the formula of G, § 59. It is believed, however, that the one now to be discussed is simpler and more useful for developing general theory. - 7. Suggestion is found in the obvious device of taking the fixed curved boundary of the obstacle as the limif of a rectilineal polygon, for which latter type of obstacle the solution of the hydrodynamical problem is, in one form or another, well known. With very different analysis Cisotti (paper 5) has dealt with a polygonal obstacle, and Villat (paper C) with a curved obstacle as the limit of a polygon. In paper D, $38, the present writer has outlined the formulation of this process in terms of curve-factors, for the particular case of symmetry, but without such demonstration of the generality of the formula as is necessary if further theory is to be built wpon it. Beginning with the asymmetrical case, it is convenient to replace (,(Z) by (—e) G(Z), so that the geometrical relation is dz = (€-c) G(€) dé. (7) Then, since £-¢ has angular range 7, what is required of ((Z), besides generality, is that it be a curve-factor of angular range zero, such that for all real values of greater than a@ or less than -@ the modulus of (2) is unity. Lest the ultimate parallelism of the free stream-lines should be regarded as an unjustified assumption, it may be recalled in passing that the angular range of a curve-factor is always 7 times the order at infinity (D, § 5), so that the constancy of | 6(Z) | for J > » is not compatible with any angular range for G(Z) other than zero. Now if in @;.(Z), as defined in equation (5), a = 8 =a, and ¢ is replaced by a real parameter « intermediate in value between —«@ and a, there results a curve-factor Ill Goal, £) = - i(nl — a?) (a? — 8) 8 + (@-0)4, (8) whose angular range is 7 and whose modulus, for @ real and & > a’, is a(a@—«?) *|C-«|. Hence 67,(«, Z)/(f —«) is a eurve-factor which, if multiplied by a suitable complex constant, satisfies the particular requirements of ((Z); and any power of this expression is equally suitable. Further, for ¢ real and @* > a’, the vector angle of the expression is constant, save for abrupt change in passing through the value ¢ =x. 18 Proceedings of the Royal Irish Academy. Tf, therefore, a succession of values x, be taken for «, all lying between —aand a, and if to ¢(Z) be assigned the form 6 (©) = KM 6:2 «r, DUE — «r)\™”, (9) where K’ is a complex constant, the relation (7) will represent upon the relevant half-plane of & the region bounded internally partly by an open rectilineal polygon having corners at x, x, etc. and c, and partly by curved lines extending to infinity; and when the field relation is employed it specifies a field of flow in which the polygon represents the obstacle, and the curved boundary the free stream-lines. Since @..(«,Z) is a pure curve factor, that is such as would if employed singly give a boundary free from corners, the corner-factors in ((2) are simply those which appear explicitly in formula (9); to these, in formula (7), must be added the corner-factor £—c. Hence the angles of the polygon, measured on the side of the moving liquid, are 27 at =c, (1-A,)w at =m, (1-A,)7 at Z=«., etc., as indicated in figure 5. It is, of course, permissible to take one of the values of « to be c, and if this be associated with a power \’ the angle at the vertex will be (2 - ’) z instead of 27. The form of that part of the boundary of the obstacle which is in dead water is, at this stage, irrelevant. An obstacle of the type represented in figure 1 may be regarded as the limit of a polygonal obstacle of the type represented in figure 5, as the sides of the polygon become indefinitely numerous and indetinitely small. If the obstacle is to be smooth everywhere save at the vertex, each angle of the type Aw becomes infinitesimal and therefore also each index A, with the sole exception of \’: such infinitesimal index may be represented by 6A. In order to get a vertical angle 2pz it is necessary to put ’ = 2p. LearHemM—On Two-Dimensional Fluid Motion. 19 The passage to limit yields the formula (ClO rin a* (ele. d)™ G(o)=K ese so Th lecaeenit y G, 2p WG = K’ pte "Taxp | log joe dX. (10) Kk=—a The integral in the second line is not, of course, definite until assumption is made of a functional relation between « and X, that is between the integrated flow up to a point of the boundary and the direction of the tangent at that point. The form of this function and the form of the fixed boundary are interdependent, so that if either is assigned the other is determinate. The corresponding geometrical relation is = _Kt- aioe G = a rae . io fel dX, (11) —a and this, together with the field relation, with the hypothesis of an adjustable functional relation between « and A, constitutes the general formulation of the hydrodynamical problem. But the generality of the formulation has yet to be proved, and this must now be done. DEMONSTRATION OF GENERALITY. 8. The corner of angle 2p7 at the vertex being assumed to be the only corner on the obstacle, though others could easily be allowed for if necessary, the theorem to be proved is this:—Any curve-factor whatever, ¢ (Z), which has the property that its modulus is constant over those parts of the real axis of € for which Z* > a@?, and which has no zeros or infinities on that axis save an infinity at ¢ of the form (¢ —c)~*?, is capable of being expressed as in formula (10). Without any real loss of generality, one may postulate further that 6 (¢) tends to a real limit, K, for ¢€—-> +4 0. It is convenient to introduce a new function defined by the relation aK G-k ee (a - I Gin (K, €) { “74 2\4) (C€ —k)j}- kK —4(@ — k*)F} y 2 ae ah see yh? > — «KG + a — i(a* — «*)2 (C* - a’)P (12) which, for « real and |x| <4, is a curve-factor in ¢ whose limit value for €—>- is unity, with constant modulus (unity) over the proper ranges of R.1.A. PROC., VOL, XXXIV, SECT, A. [4] 20 Proceedings of the Royal Inn Academy. f real. It is to be understood that the branch pomts ¢=+a are excluded by infinitesimal cavities from the relevant half plane of ¢ for which the function is defined. If dz/dZ be equal to a product of powers of functions of this type multiplied by a real positive constant, the asymptotic directions of the curve corresponding to € real will be parallel to the real axis of z. Now let a fixed value Z, be assigned to ¢, corresponding to any point in the relevant half-plane, and let a complex variable € be substituted for x. There results a function (&» - &) {= - (a? - £34) F(Z, %) = , ie ° 7 ZZ Gee i(a = ¢:)! (Gi fe, a)? (13) and if this also be considered only in the relevant half-plane of Z its definition can be cleared of ambiguity by postulating that the continuation of the function from real values of between the branch-points + @ shall he by paths confined to the positive half-plane with these branch-points excluded as before. It is important to notice that F(Z, Z,), regarded as a function of Z, has no infinity or zero in the relevant half-plane save the obvious zero at Z,, It is proposed to show that, if ((%) be any curve-factor complying with the conditions set out in the first paragraph of the present article, then G@ (Z,) can always be expressed as a product or limit-product of factors which are powers of F(x, &,), where « takes real values from — @ to a. For real values of Z it is convenient to put @(Z) =7 exp (7S). The integral az log F(Z, Z,) d log & (2), (14) taken round any contour in the relevant half-plane of f, will vanish if the contour does not surround an infinity or branch-point of the subject of integration. Consider a contour consisting mainly of a semicircle with centre at the origin and large radius /#, together with its diameter in the real axis, with a semicircular detour of infinitesimal radius + round the point c. At Z» which is assumed within this contour, log F(Z,&,) has a singularity of the type log(Z,-Z); in order to exclude this a circular cavity of infinitesimal radius «° is made round Z,, and a cut from this across to the semicircular boundary. This gives a complete contour which includes the two sides of the cut and the circumference of the cavity ; and inside this contour the subject of integration has no singularities. Limit formulae are sought for ¢ and ¢’ vanishing and R — x. It is convenient to make the cut to the negative end of the diameter, as in figure 6. LeatHemM—On Two-Dimensional Fluid Motion. 2) The logarithm is so defined that, as @ traverses the real axis from -c to o, the imaginary part of log (¢,—Z) inereases from zero to zr. If ¢ made a complete circuit round Z in the positive sense, log #(Z, 6) would increase by 2im; hence the values of the logarithm at corresponding —R rs) é R Fic. 6. points on opposite edges of the cut differ by this amount, and the integrations along the two edges combine to give 6 24 {2 log G(Z) —R or 22 {log G(Z,) — log G(- R)}, whose limit is 2i{log G(Z,) —log K}. The integral round the circumference of the cavity is of the order of <’ loge’, and so has its limit value zero. On the infinitesimal semicircle round ¢ the integral has the same limit as Lp a log F'(c, Z,) | d log (¢ - c) - namely 2p log F (c, Z,). On the semicircle of radius &, for Ry» o, log F(Z) # log; and if €=R exp 10, d@=12d0. Ti G(f)=K {1+ f(c)}, where 7(Z) > 0, then dlogG(Z) # 72 f'(Z)d@. Hence, if it be assumed that the order of smallness of f(Z), for R or | Z| great, is such that log R/(Z) > 0, the integral round the great semicircle has zero limit. ‘The asstimption is justified if f() be of the order of smallness of any negative power of ¢, as, for example, if @() be regular at infinity. The real axis contributes a line-integral in which @, being real, may be replaced by €. This integral would generally be semi-convergent as regards the infinity at ¢, but the semi-circular detour leads to the Cauchy principal value. It will therefore be understood in what follows that derived integrals which appear to be semi-convergent have their Cauchy principal values. [4*] 22 Proceedings of the Royal Irish Academy. The collection of all parts of the contour integral and the equation of their sum to zero yields the equality f== 2ijlog & (Z,) — log Kj + 2ip log F(e, Z,) + |x log F(E, Z,) d log G(E) = 0, —=-o which may also be written (15) 2i}log & (f,) -— log KH } + 2ip log F(c, Z) —E=a E=a w log F(E,Z,)dS = 0. (16) g=-o f=-0 + | log F(E, Z,)d logs +1 9. Let Z,’ be the complex conjugate to ¢,, and let a function G(Z, Z,’) be defined by the equation ; pg (2, — Z){-2 + i(a? - @)} G(z,6.) = — Se7 | aa Uae Ren ae eet — 000 + a? + t(a® — 7)? (Cy? - a’; (17) this being the form appropriate when Z is real and &* < a?; and let the function be defined for the half-plane on the positive side of the real axis of f so that the continuation from the above form shall be by paths which never cross the axis and never quite pass through the points +a. Then the form appropriate to Z real and @ > a? is (Go - 2-2 + @- a} — 22, +a? + (2? = ay (G* = a?)) G(%,%.) = (18) This function @(Z, Z,’) has important relations to the function F(Z, @,) in the particular case when Cis real. When Z is real and 2 < a’, the complex conjugate to G(Z,Z,') is F(Z, Z,). When Z is real and Z? > a’, the complex conjugate to G(Z,Z,’) is 1/F(Z, Z,). G(Z, %,") has no zeros, infinities or branchings in the relevant half-plane (&, being on the irrelevant side of the real axis), and so log G(Z,Z,’) has no branchings or infinities in the relevant half-plane. Therefore the integral \F log G (Z, Z9')d log (6) (19) must have the value zero when taken round a contour differing from that’ shown in figure 6 only by the omission of the now unnecessary cavity and cut. Just as in the case considered in the preceding article, the coutribution to the integral made by the great semicircle has zero limit, and the infinitesimal semicircle round ¢ gives 2ip log G(c,Z,). Thus the result of the contour integration, in its limit form, is [== Yip log & (c, Z,') + | log G(E, %) d log E(E) = 0. (20) =-< LearnEmM—On Two-Dimensional Fluid Motion. 93 10. The equality of formula (20) must remain true if every term in it is replaced by its conjugate complex. The conjugate complex of the G function has already been specified; that of log @(&) is logs - 73. Thus the new formula is E=a — 2ip log F'(c¢, 5) + = log #(&, Z,) (d log x — id3) =a E=-a == ie | t | {- log F(&, Z,) (d log r — idS) = 0. (21) E=-a faa Since, by hypothesis, 7 = A for & > a*, dlogr=0 over these ranges. Hence formulae (16) and (21) can be simplified, and are respectively equivalent to ; (6) noe Bic log ans p log F(e, 2) ; g=a z U aps: | tox a, 2.) dlog7 - — Ajo ner fo) ds, (22) g=-a B= . f= 0 =- plog F(e,Z,) — = [2 og &(E, 2.) dlog r gé=—-a B= Sco &=14 1 | \y _ FR 3 2): renee fe = og F(E,f)ds. (23) 2m | j \ E=—o0 f=a =-a Addition of corresponding sides of these equations gives @) i fee ~~ SAS ig A a) ia o hz 2.) ds 24 log ) |) = 2p log F'(e, Zy) =a log # (&,.) dS, 24) gé=-a so that | 2 G(f,) = K1F(6,&)| ” Exp oy et | log FYE, &) dS). (25) T - é=-a It being remembered that there is necessarily a functional relation between Sand &, this formula is recognized as an expression of ((,) in the form of the limit of a product of the character discussed in article 7, above. Thus the generality of the synthesis of 6(2,) by factors of the type of #(&,Z,) is demonstrated. ‘The form (25) is preferable to the form (10 since A is always real while &’ is complex. In the hydrodynamical application it may be convenient to take A’ to be unity. 24 Proceedings of the Royal Trish Academy: The significance of the functional relation between = and & becomes clearer if the vector angle of dz/df on the boundary is denoted by x and it is noticed that, for § > c¢, = y, while, for E 7% \ is the angle which the tangent to the boundary, drawn in the direction of & increasing, makes with the axis of x. From the relations (1) and (7) it is seen that $= all along the boundary. 11. The case of symmetrical tlow, as typified in figure 4, lends itself to similar treatment and leads to a result of the same character but simpler form. If the geometrical relation be written dz = G(f)dZ, (26) and be associated with a field relation “= Z, the problem is to find a convenient and general form for the function ((Z). What is required of G(€ in this case is that it be a curve-factor of zero angular range, giving the proper corner at = 0, and such that (i) it is real for Z real and negative, (ii) its modulus is a constant, say K, for Z real and greater than «. It can be shown that every ¢(Z) satisfying these conditions and regular at infinity can be expressed as the limit of a product of powers of f(x, &, where ta —«)* — i(Z - ayy? (a-«)'-i(Z-a)? f (x,Z) = =- * ss (27) K-¢ (a—«)24 i(€-a)* The proof of this follows so closely the method of articles 8 to 10 that only the outline need be given. An arbitrarily selected value of ¢ in the relevant half-plane is denoted by %,, and its conjugate complex by ¢,’. An auxiliary function g (¢,’) is defined by the equation (a - Zh 4 i(Z,’ - ay (a- zy! = v(Zo = a)’ (28) , ean, ZW 9(6;5) =- and it is noticed that the complex conjugate to g(é, Lgl Sens yiep) LOX — a itis 1/f(&,%,). Contours in the relevant half-plane LeatHem—On Too-Dimensional Fluid Motion. 25 of Care taken the same as in articles 8 and 9, save only that the infinity of ((£) is taken at the origin and as of the type ¢”. The integrals [= log f(, 6.) d log G(@) and |= log g (Z,f,) d log & (Z) are evaluated round their respective contours, it being known that the result in each case is zero. The results are — i log {6(£,)/K} + tp log f(0, ¢,) +| mate lag APN = © Tv = © (29) AiG LRU aes, : tp log g (0, 2.) + | = log 9 (E, fo’) (d log g + tdy) = 0. 2 Tr peR (50) —@ In the latter formula every complex is replaced by its conjugate, and it is remembered that in both formulae dy = 0 for € <0, and dlogg=0 for &>a. Then the elimination between the two results of the part of the integral which involves d log ¢ gives g=a aN ie iy a Ce a ee log {6 (60)/K } = p log (0, 2) + | — log /(E, 2), BL) b=0 which is equivalent to 1 n = 1 a= 4 = 2) P Se aN ene Pt (real : G (Ga) = Ki - — AG aa cp. ax log | = @ . AG ay at + i(& — a)?) : a ( (EF + 4% — ays) (32) the integral being definite in virtue of the fimctional relation between y and — which subsists at the fixed boundary. Formula (52) is the general formula for 6 (¢) which has been aimed at, and its generality is now demonstrated. It is to be noticed that, for € real and a@>x>0, [f(«,¢)}" has the following properties: (i) for €> a the modulus is constant, being equal to unity; (ii) for a> €>« the vector angle is constant, being m7; (ili) for €<« the vector angle is zero. Formula (32) is equivalent to formula (94) of paper D, § 38, and represents a passage from a rectilineal polygonal obstacle to a smoothly curved obstacle as a limit. 12. If attempt were made to approximate to the definite integrals of formula (25) or (32), for an assigned form of smoothly curved obstacle, by replacing the integral by a series, the resulting specification would be that of 26 Proceedings of the Royal Trish Academy. a field of flow with a number of infinities in the boundary. For the sub- stituted obstacle is polygonal, and gives rise to a field in which the velocity is infinite at every convex corner. No matter how numerous the corners, it is difficult to regard such a field as constituting an approximation to the How past a smooth obstacle. It is not, therefore, as aids to approximation that the formulae are considered; but it is hoped to show that they are useful in the exact theory.. DETERMINATION OF THE PornTs OF DEPARTURE OF FREE STREAM-LINES FROM A CURVED OBSTACLE. 13. In eases of liquid flow pasta rectilineal polygonal obstacle it is usual to take for granted that the stream-line which follows the contour of the forward part of the obstacle on either side breaks away as a free stream-line at a corner of the obstacle. But the considerations which support this assumption do not apply to a smoothly curved obstacle, and the important problem of the determination of the points of departure in such a case calls for attention. The question may be approached by considering the rate of change of the direction of the tangent to the free stream-line just at the point of departure. So far as the previous analytical formulation is concerned, any points on the obstacle may be assumed to be the points of departure of the free stream-lines. 3ut if, after such an assumption has been made, the curve of the stream-line be traced and it be found that at the very outset it enters into space occupied by the sclid obstacle, clearly the specified motion is physically impossible. Thus there suggests itself a rule to the effect that, unless the inward curvature of the free stream-line at the point of departure is less than that of the obstacle, the specified motion is impossible. This rule may, in a sense, stand; but its wording may prove misleading unless it is known and remembered that generally the free stream-line has not got a definite curvature or radius of curvature at its point of departure. This fact will be proved in the following article. For the sake both of practical utility and of keeping the main argument as free as possible from analytical complieation, it is proposed to consider in the first instance the case of symmetrical flow. 14. In dealing with a configuration of the kind typified in figure 4, every case of flow can be specified by dz = ¢(¢)dt, where @ is expressible in the form set out in formula (32), Lyaraum—On Two-Dimensional Fluid Motion. eT From this formula it is seen that, if €, have a veal value &, greater than a, and if 6(&,) = K exp (iyo), g=a 4 - ¢ 4 ~ ay? ly Be apy: = Ne — 2 tan” (Eo gu + [= Jor —2 tance {S07 a: (33) a ) a | (a _ £)}) E=0 so that t=a Xo(%) = par + | (i ebXal ye,((s)) = OW), t=0 as was to be expected. If Xo (a) + 6x9 and a +e be corresponding values of yy and &), f=a 5 Operas s 2 Fe Py aes ba (34) Xo = — 4p tan i= = =| an i) VX ( ) When « is very small the first term of this formula can readily be replaced by a simpler approximately equivalent expression; but it is not obviously legitimate to substitute {e/(a - £)\2 for tan~'{«/(a@ — £)\2 under the integral sign, since there is a part of the range of integration in which a — & is very small. If, however, «= 7’, the theorem of the mean, applied to the subject of integration regarded as a function of », gives tan {n/(a - &)3) = n(a - €)3/(a - E+ 9”), where n > 7 > 0. Thus g=a ee ane a = en afte -£|¢ &)? dx ms ox ns (3) mJ) a-€+e’ (59) z=0 where e>e>0. ‘The last integral has a definite limit value for « 5 0 (which involves «’ + 0), and therefore the equation 1 1 2 ax — dx = & | 2pa? + Z| all ite Xo T (a a E)! (50) is a valid first approximation to formula (34). Thus éy, is generally of the order of smallness of <2. Of course a higher order of smallness is possible in particular cases. If 8s be the corresponding element of are of the free stream-line, ds =| 6(€)|«=Ke. Hence generally dy,/és > ». It must be noticed that the above argument has tacitly assumed the definiteness of dy/d& for & =a. It may be taken that, throughout the present discussion, the hypothesis is that the sides of the obstacle are R.I.A. PROG., VOL. XXXIV., SECT. A. [5] 28 Proceedings of the Royal Irish Academy. smoothly curved, with definite curvature at every point except the vertex. This guarantees definiteness of d\/dE. 15. It is useful to test this result by seeing how it applies to some of the particular examples most easily available, namely @s, Gy, and 63; as defined in paper D, s§ 39 and 40. In the case of 5, it is readily seen that éy, is of the order of smallness of & loge: but then y(&)— y(a) = A(a - Ey}, and so dy/d= is not definite for E=a. It is not therefore to be expected that @, should come under the theorem of the preceding article. If @(t) = }i€.(a -— 0)/Z4{*, ther y(E)=p}a—2(E/a)#}, and dy, is found to be small of the order of « This is not contrary to the general theorem; it merely means that \(&) is such as to make the coefticient of & zero. This may be confirmed by substitution of the particular form of x (&) in formula (36). If dy/d&=-X, where X is a constant, the coefticient of ed is 2a hp - 2a\/x), which is zero if AX =pzx/2a. This is just the value of A which corresponds to ( = {¢y (a — 0)/it!)*, and this explains the fact that for this particular curve-factor 6\. is shown by the formulae of D, § 40, to be of the order of smallness of e. 16. The possibility and the importance of the vanishing of the coetticient of ed in the expansion of éy, render it desirable to carry the expansion a stage further by determining the coefficient of the term in e. To this end, consider ee s . x(&) - x(®) Ee ek (8) = X(@) og “| x(a + €) - x (@) dk, J (a - EA (a+e -€) 4 (a - &)? (a - BE) (a +e -€) (a — &)} x(a) - x(a + ®t + } + eix(@) - \(%)} eRe ME Sic A v4 fa - ry (a —~&+ «) (37) noting that the postulated definiteness of dy/d— is a security against divergence of the second integral at & = a. By the theorem of the mean the numerator equals (a —&)ely'(&’)- y' (a + €'}}, where a>&>&, and «> e’>0; and this again, by the same theorem, equals —-(a-Eje(a+e’- By y(E”), where a+e¢€ >&">£&', sothata+e>E&”>& Hence ray (EE) (a+ & - ©), He=-« penne (38) Parnes 3 LeatHem—On Two-Dimensional Fluid Motion. 29 {t is clear that generally the integral in this formula is a continuous function of « and has a definite limit for « > 0, since « > 0 involves & > 0, and € ya involves & ya. Of course the demonstration contains implicit assumptions as to the continuity of y(&) and its first and second derivatives, on both sides of the value & = a. Now a“ dé ioe y (@ + &) — x (a) > : o(a-EP(a-E + e) Jo (a= gy (w+«e-&) dé > ey (a) a It — Oe x’ (a) [tan (u - tay] ? 0 Dey’ (a) {4m - tan(«/a)*}, > my (aye. (39) On substitution of this in H it is seen that (48) is equivalent to “ By) ra v(&) — & 1 ee 2) OE ig = | MOESRO re on) (Co) Jo (a - Era — = + €) m7 Gre ey’ f where w is small of the order of e. Now if the integral in formula (34) be integrated by parts, it being remembered that (0) = pz, there results T — dy, = aa (a) tan? sn dl (41) XO hae = Ney Mie ) ) \ and therefore, by (40), when ¢ is small i 2 1 — dx) = & i a4y(a) — — Tv (iis WS) Gy, OG a | - eta) (42) Jo (a —&)? small quantities of the order & or higher being neglected. Integration by parts within the square brackets leads to ny , A 2p 2 |e dé ] , ) 43) NG CN a Be rs PN eareyny eee (43) This is the desired improved approximation. It will be convenient to denote the coefficient of < in this formula by S. A very important inference from equation (45) is that, if the coefficient of « be zero, — OXo equals the angle of contingence of the fixed boundary for an element of are Ae. Thus in this case the free stream-line and the fixed boundary osculate at their point of separation. [s*? 30 Proceedings of the Royal Trish Aeademy. 17. When the sides of the obstacle are convex to the liquid, dy is negative, and it is clear that if a be small the ratio of the second term to the first in the coefficient of & is numerically small. Thus if the point P of departure of the free stream-line be supposed to take positions successively further and further from the vertex of the obstacle, - dyo = eS, where S begins by being positive but keeps getting smaller. So long as S is positive the free stream- line is curving sharply inwards into the obstacle, and so the motion is physically impossible. But when a@ is so chosen that S = 0, dy. is small of the order of «, and the free stream-line osculates the curve of the obstacle. The point of departure of the free stream-line in this case may be called P,. Without closer consideration of the form of the obstacle behind P, it is not safe to say that a free stream-line can actually depart from P, and be clear of the obstacle ; but, so far as the shape between the vertex and P, is concerned, P,, determined by S=0, may be described as the most forward point of the obstacle from which a free stream-line can break away. If a greater value be assigned to a than that corresponding to S = 0, generally S may be expected to have become negative, so that - Sy, is of the order of ¢! and is negative. This would indicate a free stream-line breaking away with a sharp concavity to the moving liquid. It may be thought that this is physically less probable than the smooth departure at P,; but such an opinion is somewhat speculative. Anyhow the sharp outward turn gives increased chance of clearing the hinder part of the obstacle, so that if departure at P, were impossible it might well be possible from a point very near to P, and behind it. The theoretical importance of the point P, is in any case obvious, and it is to be remembered that the corresponding value of a is determined by the relation at = J (a - E)? j (44) 18. It is interesting to note that there is available one rather compre- hensive type of curve-factor into which the parameter a enters in such a way that the condition S = 0 is satisfied. This corresponds to (sy of D, § 40, and may be written » L . G3 (€) = Exp. log Sa ska ix! f (x) dx, { fe are 0 where f(x) is an arbitrary function. LeatarmM—On Two-Dimensional Fluid Motion. ra | For a real value & of ¢, intermediate between 0 and a, the vector angle Xx of this curve-factor is a ge x (é) = Saf («) de + | tan'{«/(E - x) FEF" («) de, 4 a E = i0 Ir («) dk - | tann4¢e a Kk} f’(«) dx ; (46) 0 and from this pr = x(0) = ar} f(a) —f£(9)}, (47) dé | 2 es oe EEE a ax al eo oi (48) On substitution of this value of dy t= Fi f ay (ide Fe &)} (B=0% ie lie =0 ES Though the subject of the double integration has an infinity for «x = 0, & = 0, this is not of sufficiently high order to preclude change of the order of integration. If this change be effected it appears that t=a K=a t= 0 i GK, sea lig Aide L ims 2 i = 8)? E(a@ — &)°( - x) é=0 k=0 t=k ; Deere an Ve = [aes (x) ($5) dk, c=0 =— Lea?) f(a) — f(0)) =- pra’. (49) This shows that for all curve-factors of this particular class S = 0. Formula (45) should be compared with formula (4) of article 6, above. It seems that formula (4) cannot be general, since it gives a curve-factor possessed of this very special property. This casts doubt likewise on the generality of formula (5), and suggests the possibility that Mr. Levy’s method of approximation (paper E) may be less general than it seems, 32 Proceedings of the Royal Irish Academy. Pornts OF DEPARTURE OF FREE STREAM-LINES IN ASYMMETRICAL FLow. 19. It has been shown above (article 10) that every case of asymmetrical flow can be dealt with by the relation dz 7 - _2 | 1 # + zz = ElG-MM EL) *Exp|-—} |" + [7 | log F (E,&) dy |, « \ where (50) P(E.) = (& - &) 16 + t(a? — Bee, - a? + i(@ - EAE? - a), (51) and x is afunction of &. It follows that if y, be the vector-angle of d2/dZ, corresponding to a real value &, of &, and if &,* > a*,’ (ae eee - et Bee xo) = [=] - 2p ) cos? : cos 15 - = + r | [ | feos 4 — — cos? ae = (52) the term in the sonare paaee Wes omitted when —,> a, but retained when £&, < - a. Thus B e 2 [em $ = are =.9 a) peek es EES Xda) — yo @ +) = 2p cos Reser. = 1 PD ha Sac dy: nee | la(a — E) + ans 1 ! —a ¢ If the theorem of the mean be applied to the subject of integration, regarded as a function of » whose value is zero for » zero, the integral is seen to be equivalent to ra p @-éjhdy Je ) (2a + wy?) (a - & +7)?” . @a3 eal where n>» >0. This latter integral tends to a definite limit for » — 0, namely the form got by omitting »’. Also, when 7 1s small, the first term of (53) approximates to {2 (a+c)/a.a—c)j'y. So, if 4° =e and is very small, formula (55) yields the approximation He! tetas clay ioaeson Xe(@) — y.(@+ e)-e€ ftProer eAdf af-)[2ZG2 OV gles es at (| ilaq-p} el Leararm— On Two-Dimensional Fluid Motion. 33 Similarly $f 2(a—-c))4 ead) = BN (— = » —— RSME OL MM Dyes ez er o| Dip ize 2 + ife] Haeret a] The same argument may be applied to this pair of formulae as has been employed in the case of a symmetrical obstacle; when the form of x is assigned the simultaneous equations got by equating both coefficients of e to zero determine values of @ and ¢ corresponding to the flow in which the two points of departure of free stream-lines are as far forward as is physically possible. 20. The approximation to y(@) — xo(@ + <), When < is small and positive may be carried a stage further as follows. Since d ee | i (2a + ©)? dé ! ala — E + «) (a — £2)3 (a — & +e) and since y(¢ -) — y(¢+)=(1 = 2p)m, it appears, on integration by parts, that No(@) — xo(@ + «) ff 2) — @ 1 st re wr ry, = vos! eae =| = tee 4(2a + €)? | + | rz ae x(%) -ale )(@ -— &)2 (a - E+ «) (56) Also, by repeated use of the theorem of the mean, xX@ + «) — x(a) dé i’ = \ x(&) = x@) ee lee x() oe : (a - E23 (a = E + ¢) bed re _ #2 Ae & | (a ae E)3 (a - Fa +«) - [et ){x(a) - sisal x(&)} 5. Soren Ge (a? -— & -E)(a-& | eo PGA) (Oe) (GE ee) \"( ig? ies (a+ & — &) dé =-€ (@— &)8 (a - E+) ; (57) where a@>&>& ese >0, at+e>€&’ >, and the integrations are from —a to a. The last integral, if \’(&) is assumed continuous, has a definite limit for «50; for if e50 then also & 50, andif & 5a then also & ya. Hence //’ is small of the order of e. 34 Proceedings of the Royal Trish Academy. The third integral in H’ depends on a Pees edb bee eel le=anGers A Qa +o} [es la(a—E4e)$j_. Tv ~ dQa+ oh (58) Therefore the smallness of H’ makes formula (56) equivalent, when & is neglected, to Xo(@) — Xo(@ + €) Sip ot til nd ie x (E) - x(@) (a(a—c+e)) = ~A(a+e EK (a— Tae x(a + €) + x(a). a (59) of este [o-n0o| CVT LTH fa o {l3 a+e\t el 2p) x (<=) and therefore (59), when & is neglected, reduces to 2(a+c))4 1(f 2(a +&) eral *rl\*|}laerey &x|a-x@e -ae ) _ This shows that, when the parameters are such that the coefficient of e} is zero, the angle of contingence of the free stream-line at its point of departure is the same as that of the obstacle, so that the two curves osculate. a = COs a Xo(@) — xo(a + €) = [2p INFLUENCE OF THE SHAPE OF THE OBSTACLE UPON THE DIVERGENCE OF THE FREE STREAM-LINES, AND THE RESISTANCE TO RELATIVE FLow. 21. The theoretical considerations of the present paper have precise bearing only upon flow which is everywhere irrotational, with discontinuity at the free stream-lines. This may, for the moment, be called the “theoretical tlow”; it is not realized in actual fact or experiment. What is realized seems generally to be a continuous flow which is not wholly irrotational, and this may be called the “actual flow.” In the theoretical flow there is a dead-water or wake extending to infinity; in the actual flow there is a wake of limited extent within which the motion of the fluid is rotational. LearHeM—On Two-Dimensional Fluid Motion. 35 For the same obstacle and the same limit velocity at infinity, the theoretical and the actual flow may be said to correspond. It is doubtful whether a good case has been made out* for believing that the resistance of the obstacle to the flow is the same in corresponding motions; nevertheless the correspondence may help towards estimating the resistance in the actual flow. It seems to be established as an experimental fact that in actual flow the resistance is greater when the wake or region of rotational motion is extensive than when it is small. Thus, in such a problem as the designing of the cross-section of a strut for an aeroplane, attempt is made to secure a form such that the lateral spreading of the wake, and its area, shall be as restricted as. possible. Now it may be argued with a certain degree of probability that, the less divergent are the free stream-lines at their points of departure in the theoretical flow, the less spreading will be the wake in the actual flow, and therefore the less will be the actual resistance. For different shapes of prow the capacity to produce divergence of the two branches of the divided stream is likely so to correspond in the theoretical and the actual flow that where it is relatively small in one it is also relatively small in the other. In both kinds of flow divergence of the wake is evidence of the capacity of the prow to screen the region behind it from the force of the adverse stream. If this be true it is important to enquire how, in theoretical flow, the divergence of the free stream-lines can be made as small as possible. And it may be surmised that, in theoretical flow, the most probable points of departure of the free stream-lines are those points which (as discussed in articles 17 and 18, above) are the most forward points of the obstacle from which departure is possible; or, at least, that the degree of divergence for these most forward points of departure is relevant for purpose of comparison with the corresponding actual flow. When the flow is symmetrical the most forward possible points of departure correspond to the value of @ which satisties the equation S=0, that is (61) where there is a functional relation between y and & whose form determines or is determined by the shape of the obstacle. Assuming p to be assigned, the question of design with a view to minimizing the divergence of the wake * See a foot-note near the beginning of paper A. R.I.A. PROC., VOL. XXXIV, SECT, A. (6) 36 Proceedings of the Royal Irish Academy. is simply this :—What kind of relation between y and & will make the negative range of integration with respect to y, that is from y=pa to x = (a), as great as possible numerically, so that y(@) may be as small as possible ? ‘When the curved sides of the obstacle are convex to the stream dy 1s negative throughout the range from —€=0 to =a. Andas ai/(a—£)? is greater than unity, the whole range must be numerically less than pz, so that y(«) is necessarily positive. Thus the stream-lines must diverge; but the divergence may be kept small. Tf formula (61) be written Tt — Lim Sf at/(a = Ey (- ox), (62) it is clear that if the greatest values of — dy be associated with the smallest values of (ad/(a ~ E)h) they contribute less to the sum, and therefore a greater range of — dy is required to bring the sum up to an assigned value. So the association of the greater values of - éy with the smaller values of & diminishes the divergence of the wake. In other words, if the rate of turning of the tangent to the curved side of the obstacle be greatest near the prow, this configuration makes for reduction of the divergence of the wake. 22. It will be noticed that in the last sentence, where the word “curvature” suggests itself, a different phrase has been employed. This is because the argument has been founded on a functional relation between x and E, and the “rate of turning of the tangent” which appears in the result is not — dy/ds, where s is the are of the contour, but —dy/d&. If o is the curvature of the obstacle, a = — dy/ds = — (dy/dé)q, where q is the resultant velocity ; and the theorem cannot be expressed in a purely geometrical form until it is ascertained how change in the value of - dy/dé affects the value of g. It cannot be assumed & priori that - dy/dé and o increase or decrease together. It will be advantageous to write dy’ for —dy, remembering that this is always positive on the convex obstacle. If an addition 8y’ be made to dy’, and if this be concentrated at a single point £ = , of the contour of the obstacle, a corner is created there, and g becomes either zero or infinite. It is therefore necessary to suppose Sy’ to be spread over a range of & surrounding the point &,, say from —&-—y to &+-y; if 2y = 8, it will be assumed that oy’ and o& are infinitesimals. If & = £ +A, the addition to dy’ over the postulated range may be represented by @6(A), where @ is a function such that ay =|" d G(X) = O(y)-O(-y). (63) LearaemM—On Two-Dimensional Fluid Motion. oF For & real and a> &é,>0, gé=a 2 3) p , EG 2° } G(&) = KS Seay Exp [x log ie ee = 2 ait (S(O e)) sy) 7 @=tPa=e) é=0 and therefore, at &, since g = | 1/¢(&,) |, b=a (ad —£,)#)? dy’ = a - (a- -logg=1 | ie + (a= 50)" | [2 aa an TEA (64 84 “SL la - (a- Ea) is F —€)84 (a-&,) ) t= If dg be the change in qg due to the dy’ above specified, it follows from formula ae that r 4 (a-&,) (w= £)-d)} - (@-£, (a-& - r)? + (a —- = om 7 Sy eae TOCA) ae a si heal at the last ae oo approximate, the second power of A being neglected under the logarithmic sign. Thus Y 7 pee a (ae(inE zy a | leelegey. Ga Gf 7 T =7 =% in which it is to be noticed that the first integral, by (63), equals éy’, and that the infinity at \ = 0 in the second integral is not sufficiently powerful to interfere with convergence. It may be assumed that @(A) is expansible in the form 4+ 5+ CX +..... ; so that d0=(B + 2CX +..... ) dX; then % Y | log || a0 =[(2 MOON op WAU = 2B les y =%) =F approximately. Also, by (63), 2yB=éy’ approximately, Therefore the second term of (65) equals (dy’/7) (log y-1), and = 2 & Ieaeiaieleatiaioe Gy oul 66) ; 7 [tos ( £) g {4 (a - &)} IF ( This formula represents the change in g at a particular point due solely toa variation éy’ distributed in the immediate neighbourhood of that point. If account has to be taken at €=£&, of a variation of dy’ elsewhere, say Ox’ at § =&, there is a further variation of g, namely given by eae rp Se 4 31 eens, } = Og 2 ais: log (GS as a) (67) q Tr (a — &)? + (a - &)? 38 Proceedings of the Royal Irish Academy. As regards formula (66), what is contemplated is a variation of the rate of turning of the tangent at and about &,, and if the mean of this variation over the range é& be é (dy’/d&), then éy’ = 6 (dy’/dé) d&. Hence (66) is equivalent to &@_ 1,/d& \a - 4 = 23H) [he ie 5 1 (68) Now, in spite of the greatness of logéf, the product 6s& logéE>0 as 6£> 0, and thus the formula shows that ég tends to vanishing smallness in comparison with- 6 (d,'/d5). A corresponding result holds for é,9. Thus it appears that infinitesimal variations of the rate of turning of the tangent, of the kind contemplated, corresponding namely to alterations of curvature which are infinitesimal and do not introduce corners, may be regarded as leaving the velocity g unaffected. And therefore a variation which increases dy’/d— increases the curvature, and conversely. It is accordingly legitimate now to express in geometrical form the principle obtained in the preceding article, namely in the statement that the more the curvature of the sides of the obstacle is brought into the neighbourhood of the prow the less will be the divergence of the wake. Of course complete concentration of curvature at the prow must be ruled out, for that would reduce the obstacle to a plane lamina edge-on to the stream. Other considerations than resistance to the flow have to be taken account of, and both p and the degree of concentration of curvature near the prow must be chosen so as to give the obstacle as much breadth as it requires for strength or for any other purpose. But the present principle, if it is sound, furnishes a method of bringing the resistance question simply into the balance with the other relevant considerations. 23. The same principle may be established for an asymmetrical obstacle by considering the formulae got by equating to zero the coefficients of <4 in formulae (54) and (55). Since dy is negative for curves convex to the region of flow. it will save confusion to put - dy=dy’. And the two integral signs may be replaced by a single one covering the whole range, provided it be agreed that the sudden change in the value of y at =c is excluded from the integration. The equations determining the most forward points of departure may then be written f=s f/f +=. 4 n= | (228 ay = sen (22 5) a t=a fr /a-E,, i he | (Sg) x - eS (70) §=-« LearHem—On Two-Dimensional Fluid Motion. - 39 Though there is no loss of generality in dealing with an arbitrarily selected value of a, it is not permissible for the present purpose to keep the value of c fixed. Soc must be eliminated. This gives tL } US : Tho pS | (<25) ax | (<4) dy = Apr. (71) —«a —a If, for one selected value of &, a small increment dy’ were made in dy’, the increment of Z,.2, would be /(&) dy’, where pont at ea ee ; is =) + ( a — an a-c a+ &, Hence if two simultaneous increments of dy’, namely d.y’ for & = &, and 6.’ for &=&, be such that they leave the equality (71) still true, it is li L Ss necessary that F (é) Oy’ + f (&) Coy = (1), (73) Now, the two terms in /(&) have a product independent of & and are equal when &=c. Hence /(&) has its minimum value for € =¢, and increases as & varies from ¢ towards either @ or —a. And so, if a>&>,>e, or ‘¢>&>&>-a, then /(&)>/(&), and therefore, by (73), | dy’ | > | dy’ |. Thus if 6,y’ is positive duy’ is negative, and 6,y’ + é.y’ is positive, so that there is a net increase in the whole range of y’. This means that any change in the functional form of y which has the effect of bringing curvature nearer to the prow ¢ increases the negative range of y, and so makes the free stream-lines less divergent at their points of departure. The bearing of this principle on the resistance in a case of actual (as distinguished from theoretical) flow is the same as when the flow is symmetrical. R,I-A, PROC., VOL, XXXIV, SECT. A, [7 eto] IIL. ELECTRIFICATION BY FRICTION. By PROFESSOR J. A. McCLELLAND, DSc., F.BS., AND REV. C. J. POWER, S.J., M.Sc. Read Fesuvany 11. Published Ari 17, 1918. Tue work described in this paper may be looked upon as an attempt to study frictional electricity in an accurate and quantitative manner. It has been found possible to overcome the well-known experimental difficulties of the subject to a sufficient extent to enable consistent results to be obtained. We have studied the effects of different pressures between the rubbing surfaces, different relative motions, changes of temperature and of the amount of moisture in the air, and finally we have carried out experiments in gases at greatly reduced pressures. None of these subjects has been examined fully; we have in the first place made a general survey of some of the different factors which influence the rate of production of frictional electricity, with the object of deciding on the points most likely to repay fuller investigation. The work described in this paper was carried out two years ago, but pressure of other work caused an interruption, and it was only recently that it has been possible to take up the detailed study of some of the points touched on in this paper. Work on frictional electricity has been published in recent years by Morris Owen,' Morris Jones, French? and Shaw,‘ but there is not much direct overlapping in the case of our work with that of these authors. APPARATUS. After several attempts along different lines the following method of producing friction between surfaces was found to give satisfactory results. 1 Phil. Mag., April, 1909. * Phil. Mag., Feb., 1915. 3 French, Physical Review, vol. ix, No. 2, Feb. 1917. 4 Shaw, P. E., Proc. Phys. Soc., vol. xxvii, April, 1915; Proc. Roy. Soc., vol. xciv, 1917. M‘Cieutiann and Power—Electrification by Friction. 41 The apparatus was quite simple. A circular disc of wood was hollowed out on one face to a depth of about a quarter of an inch, leaving a circular rim around the edge. Silk—or linen—was stretched over the hollowed out face and bound firmly in position by wires fitting into shallow grooves on the circumference of the disc. The dise was fixed to an axis that could be revolved by a motor so that the silk rotated in its own plane and formed one of the rubbing surfaces. Discs of different sizes up to 14 inches diameter were used at different times, and a wide range of speed of rotation was avail- able. The other rubbing surface was a metal, and different metals have been used. Small metal spherical caps of large radius of curvature were used. They could readily be attached to a simple lever arrangement that allowed them to be pressed against the silk surface with any desired pressure. The stem carrying the metal surface was insulated and joined to one pair of quadrants of an electrometer. The rate of production of charge on the metal surface was measured, either by observing the rate of charging of the electro- meter, with a suitable capacity attached, when the motion of the silk surface was steady, or more often by connecting the metal to earth through a high resistance, and using the electrometer to measure the steady potential difference between the ends of the resistance. This apparatus was modified for some of the experiments and especially for the work in gases at reduced pressures, but the general method remained the same. It will be observed that in our method of working the metal cap which forms one of the rubbing surfaces is always either at zero potential or differs from it by only a fraction of a volt. On the other hand, the silk or linen which forms the other rubbing surface retains its charge to some extent, and we do not know the potential which it reaches. After some revolutions of the disc, that is in a very short time, the steady condition is no doubt reached when the leakage from the silk balances the rate of production, and it is the rate of charging of the metal surface under these conditions which we measure. No doubt there is uncertainty as to the extent to which the charging of the metal surface is influenced by coming in contact with silk electrified during the previous revolution of the disc, but we have used this method— at any rate up to the present—because of the steadiness of the results obtained by it. Besides, the difficulty referred to cannot be avoided completely by any method of experimenting. When two surfaces are electritied by friction, recombination of the charges will always take place at a rate depending on the conductivity of the bodies and on the relative velocity at the point of contact. (6*] 42 Proceedings of the Royal Trish Academy. Variation of rate of production of electricity with speed of moving surface, and. with pressure between surfaces. The first point investigated with the apparatus just described was the effect of the speed of the moving surface on the rate of production of electricity. Low speeds were first investigated, the metal cap being fixed so that it was rubbed by a length of 6 inches of silk in each revolution. The results with platinum, iron, and silver, rubbing on silk and getting in each ease a negative charge, are shown on fig. 1. “T BGeae Metals(-) on Silk 3 Rate of Production of Charge 200 3 TA A A : N° of Revolutions of Disc per sec (/ Rev"=6 inches) EE ere EE fe ee a a ° 2 ¢ é 3 70 12 ‘4 16 18 20 FLL. It is clear from these curves and from others not reproduced in the paper, that for small speeds the rate of production is proportional to the excess of speed above a certain small value which differs for different metals. The relative rates of production for different metals under the same pressure would therefore differ according to the speed of the rubbing surface. The maximum speed represented on the curves in fig. 1 is only 12 inches per second. Higher speeds were then dealt with. The circumference of the circle which the metal cap travelled over in one revolution was now 2 feet, and 28 revolu- tions per second could be obtained from the motor, giving a maximum speed of the rubbing surface of 56 feet per second. In stating the results we must distinguish between cases where the metal got a positive charge and cases M‘Creti ann anp Power—Lleetrification by Friction. 43 where the charge on the metal was negative, for the variation of production with speed is different in the two cases. As an example of the results when the metal is positive we give curves (fig. 2) for copper rubbing on silk, the different curves corresponding to different pressures between the copper and the silk. Somewhat similar curves were plotted for zine and iron, the charge on the metal being positive in both eases. FaliGrrer /20 Copper (+) on Silk. Pressure = /20 gfams 100 Pressure = 99 grams 80 Pressure = 60 Grams sales 2 60 Rate of Production of Charge Pressure = 30 grams 20 N° of Revolutions of Disc per Sec. (/ Rev7=2 feet) Se See oO 26 sé ar M2 /#0 468 13-6 22-4 252 280 As the speed of rubbing inereases, the production increases to a certain point; a maximum rate of production is reached which is not altered by a further large increase of the speed. The maximum rate of production of a positive charge on copper was reached at a speed of about 20 feet per second, and no further increase of production took place when the speed was altered to 56 feet per second. The speed required to give the maximum rate of pro- duction of positive charge on zinc or iron was greater than in the case of copper, being about 30 feet per second. The effects of different pressures between the metal and the silk are also shown by the curves on fig. 2, the production of electricity increasing with an increase of pressure whatever may be the speed of the rubbing surface. The maximum rates of production at different pressures are roughly proportional to the pressures. In similar experiments with metal caps which got a negative charge, the results were less certain. Sometimes numbers would be measured which showed that the rate of production tended towards a maximum at high speed, and in other cases the rate of production increased even more rapidly than the speed. 44 Proceedings of the Royal Irish Academy. We cannot fully explain the want of steadiness in the numbers for negative charge as compared with those for positive charge, but there is one factor which should make a decided difference in the form of the curves connecting production and speed according as the metal surface is positive or negative. Later work described below showed that an increase of tempe- rature of the metal cap decreased the rate of charging, provided the charge was positive, but increased the rate of production of a negative charge on the metal. An increase of temperature tends to make the metal surface acquire a negative charge. As an increased speed of rubbing means an increase of temperature, this efiect would tend to flatten the curve and cause the pro- duction to reach a maximum when the metal was positive, while it would tend to cause a more and more rapid production at high speeds when the charge was negative. In addition to this temperature effect we must remember that at high speeds the rapid return of the metal over the same silk surface brings it into contact with an oppositely electrified surface. This should flatten the curves connecting the rate of charging with the speed whether the charge on the metal is positive or negative. When the metal is charging positively, we have, therefore, two effects acting in the same direction and diminishing the observed rate of production of charge at high speeds, but acting in opposite directions when the metal is charging negatively. This reasoning does not, however, explain the variable nature of the results we obtained when the metal was charging negatively. These measure- ments were among the first made, and before we had learned some of the conditions necessary for very constant results. To obtain a positive charge on the metal when rubbed with silk it is well to have the metal surface polished. When polished, some metals give a positive charge and behave very steadily. A clean surface of the same metal free from polish may charge negatively, and clearly inconsistent results may be obtained at a transition stage. To return to the results with the metal surface charging positively, the curves would require correction for both the effects mentioned above in order to arrive at the true relation between rate of production and speed, but there is no doubt that the curves tend to a maximum as the speed is increased. To obtain similar curves for negative charges on the metal it will be necessary to prepare the surface in a standard way without actually polishing it. We have tried some direct experiments with metal surfaces polished and then rubbed very slightly with emery paper. The result in most cases was that contact with the emery paper tended to make the metal charge nega- tively. If charging positively when polished, it gave either a smaller positive M‘CLELLAND AND Powrr—Electrification by Friction. 45 charge or else a negative charge after the use of the emery paper. It is probable, we think, that when sufficient care is taken to keep the nature of the surface constant during the experiment, the curves for metals taking a negative charge will be of the same type as those given in fig. 2 when the metal is charging positively. Variation of production of charge with temperature, speed of rubbing and pressure remaining constant. We have tried a few preliminary experiments on how the rate of produc- tion of charge depended on the temperature of the metal surface. For this purpose we used a hollow metal cap with two tubes attached so that a stream of water could be passed through it. The vessels from and to which the water flowed were insulated, so that the rate of charging could be measured with or without the flow of water. Care was, of course, taken that the stream of water did not of itself produce any electrical effect. Small junctions soldered to the metal enabled the temperature of the rubbing surface to be estimated. By placing junctions close to and at greater distances from the rubbing surface we decided that the temperature of the surface in contact with the silk certainly did not reach 100° C. at the highest speeds used. We could assume, therefore, that the metal surface in contact with the silk was raised in temperature by passing steam through the cap, and lowered in temperature by passing a stream of water at the temperature of the room. The results were similar in all the cases we examined. When the metal was charging positively, the stream of cold water increased the rate of charging, while the effect of steam was to decrease the rate or even to change the sign to negative. When the metal was charging negatively, the steam inereased the rate of charging and the cold water decreased it or changed the sign to positive. It is important to note that an increase of temperature produces the same effect as a slight roughening of the metal surface. When the metal surface is very slightly rough, the actual points of contact may be considerably raised in temperature, although the average temperature is not appreciably altered. The effects of changes of temperature will obviously be worth full inves- tigation. Variation of rate of production of charge with amount of moisture in the air. To test how the rate of production of charge varied with the humidity of ~ the surrounding air, the apparatus was slightly modified so that the humidity could be altered as desired. An extension of the axis of the motor, working 46 Proceedings of the Royal Irish Academy. in a closely fitting tube, projected through the side of a box and carried the rotating disc, which was thus mounted with all its attachments inside the box. One side of the box could be removed to adjust the apparatus, and when replaced could quickly be made sufficiently air-tight. The necessary electrical connexions passed through insulating plugs in another side of the box. To get observations at any humidity the method followed was to dry the air inside the box in the first place by phosphorus pentoxide, then shut off the vessel containing the phosphorus pentoxide, and start a slow stream of air saturated with moisture into the box, an exit for the air being opened. The air inside the box was thus slowly supplied with water-vapour, and during the change in humidity the observations were continued. Knowing the volume of the box, and the amount of air entering and leaving it per second, it is easy to calculate the humidity at any instant, assuming that complete mixing takes place, and this can be facilitated by means of the rotating dise. Let the volume of the box be Vec.cs., and the volume of air entering and leaving it per second Vc.cs. Let p denote the water-vapour per cc. in the box at any instant, and / the water-vapour per c.c. of the saturated air entering the box. We have, therefore, Lane pent ap (PN) = VP - Pd, which gives Humidity = i Es (ies Bie In practice we plotted the rate of production of charge against the time elapsed since the flow of saturated air into the box started. The humidity at any point of the curve can be obtained by substituting the corresponding value of ¢ in the above expression. Having first thoroughly dried the air in the enclosure, saturated air was then passed in and curves taken with various metal caps, some of which charged negatively and some positively. Fig. 3 is typical of those obtained when the metal knob charged positively. As moisture was gradually added the rate of production increased up to a certain point, and then decreased rapidly when more moisture was added. When the metal cap charged negatively the form of the curve was differ- ent. Fig. 4 is typical of those obtained in such a case, The surface rubbing M‘CxeELLaNnpd AnD Powrr—Electrification by Friction. 47 against the metal was in these experiments usually linen, but silk was some- times used. There is thus a marked difference in the form of the curves showing the variation of the rate of production of electrification with amount of moisture a S FIG. 3. lron(+) on Linen vt a is) Humidity=-e ~ g where =1000 (approx | a is} kate of Production of Charge Ny 8 s Time of Flow of Air in Mins. o 10 20 30 40 50 60 70 present according as the metal surface is charging positively or negatively. There is one factor which we must remember in considering these curyes,fand which explains some of their features, but is not sufficient to account forjthe /20 3S iS 10048 S $ 801.8 8 FIG. 4. 3 . : \e Nickel (-)on Linen. 8 Humidity=1-e m sais where Xx = 848 20 (ie Time of Flow of Air in mins. Oo 10 20 30 40 50 60 70 80 30 j00 a fez difference between the two types. As the amount of moisture increases, the conductivity of the silk or linen also increases. The initial small rate of R.I.A. PROC., VOL. XXXIV, SECT. A. [8] 48 Proceedings of the Royal Irish Academy. production when the air is very dry is accounted for by the high insulation of the linen which thus gets charged oppositely to the metal, and this opposite charge leaks back to the metal on successive contacts. We tried the effect of placing a large number of sharp earth-connected points close to the surface of the linen, and found that when the metal was positive little effect was pro- duced on the form of the curve, but when the metal was charging negatively the presence of the points increased its rate of charging in dry air. This is in agreement with the fact that the point discharge takes place more easily when the point is negative. The small apparent rate of production of charge in a very dry atmosphere may therefore be explained by the high insulation of the linen or silk. The varying conductivity of the linen as the amount of moisture alters does not, however, explain the essential difference in the two types of curves. It does not explain why the apparent rate of production as we approach saturation falls off rapidly when the metal is positive and not when the metal is negative. Experiments in Gases at reduced pressures. We sow proceeded to carry out experiments in air and other gases at reduced pressures. A more elaborate apparatus was constructed, retaining, however, the same method of producing the friction and measuring the rate of production of charge. The closed space was formed by a bell-jar resting on a metal plate. An iron tube passed through a hole in the metal plate and projected below it to a length greater than the barometric height. The disc on which the silk or linen rubber was stretched, was attached to an iron rod which passed down through the barometer tube and was rigidly fixed to a suitably shaped mercury reservoir, which could be rotated by a motor. It was easy to arrange the reservoir so that no mercury was ejected even at high speeds. The necessary connexions were made through insulating plugs passing through the metal plate. With this apparatus the necessary adjustments could be made, the bell-jar put in position, made air-tight with wax, and a pressure of ‘001 mm. or lower reached by a rotating mereury pump in a short time. Liquid air was not available, and with an apparatus of this type it was not possible to reach much lower pressures without too much waste of time. The usual method of working was to reduce the pressure as low as possible, then set the motor working and take observations admitting gas in small quantities at intervals. A two way stop-cock allowed small quantities to be admitted when desired. The pressure was measured with a M‘Leod gauge. M‘Cuvtianp and Powrr—Llectrifieation by Friction. 49 Linen was used as the rubber, and curves were plotted for a number of metals all taking a negative charge. Fig. 5 refers to platinum, and the other curves were quite similar. At the lowest pressures reached the rate of pro- duction of charge was very small. As the pressure increased, the rate of production rose quickly to a maximum value which in all cases occurred at a pressure of about ‘08 mm. of mercury. The rate of production decreased as the pressure was further increased and became more or less steady about *) mm. 70) 60 FIG.5. Platinum(—)on Linen. 304 Rate of Production of Charge. Pressure in mms. of Mercury. ak - == _— S i) 7 2 3 a 5 “6 “7 7) FEL The remarkable feature of all the curves was the regularity with which the maximuin occurred about the same pressure. Similar experiments were next made using silk as the rubber. Fig. 6 refers to nickel which is again charging negatively. The curve shows two maxima, one at a pressure of ‘(08 mm. and the other at 03 mm. We obtained the same type of curve showing the same maxima for platinum. Using a different sample of silk, the maximum at ‘08 mm. disappeared, and that at "03 mm. remained. The obvious suggestion is that the first sample of silk was not pure. We now endeavoured to plot similar curves when the metal was charging positively. As mentioned frequently above, it was easy to get a well-polished metal surface charging positively with the rubbers we used; but in these experiments in air at low pressures it was only in the case of copper or iron rubbed with linen that we were able to get a positive charge on the metal. When these metals were charging positively, an increase in the rate of pro- duction took place as the pressure was increased from very low values, but 50 Proceedings of the Royal Irish Academy. the rise was slower than in the case when the metal charged negatively, and the maximum occurred at a higher pressure. The pressure corresponding to maximum charging was also different for the two metals, being about ‘8 mm. for copper and 1:25 mm. for iron. From all these results it would appear that the pressure at which maximum production occurs is determined by the substance which is charging positively ; thus different metals taking a nega- tive charge gave the same point of maximum production when the same rubber was used, but a different point with another rubber. Again, copper and iron gave a maximum at different pressures when taking a positive charge. 240 1 He FIG. 6. Nickel (-) on Silk [a 2S ——— a Rate of Production of Chari 4 l FPressure_in_mms.of Mercury ° 7 2 3 ’ s 6 “J fEL In all the above experiments at low pressures the air was well dried with phosphorus pentoxide, and the effects of drying previously recorded must be borne in mind in interpreting the present results. The rubber no doubt insulates well when well dried and therefore retains the charge of opposite sign to that on the metal, and if for any reason the rubber lost its charge more freely at a particular pressure, the form of the curves might be explained. But we do not think that a sufficient explanation can be found along those lines, and it would therefore appear that the actual rate of production of charge varies with the residual pressure of the gas as shown on the curves. This pressure effect is under further examination in the laboratory. We do not propose at this stage to discuss the theory of the results contained in this paper. Indeed, the main object in publishing the results in their present form is to show that it is quite possible to make measure- ments on frictional electricity with such a degree of accuracy and consistency as to justify a complete study of the subject. IV. THE IONS PRODUCED BY BUBBLING AIR THROUGH ALCOHOL, By J. A. McCLELLAND, D.Sc., F.RS., AND P. J. NOLAN, M-Sc., University College, Dublin. Read Aprit 8. Published Aveusr 7, 1918. IN a previous paper' the nature of the ions produced by bubbling air through mercury was examined, Several different groups of ions were found, and with the view of obtaining further information on the nature of these ions we have undertaken similar experiments on the ionisation produced by bubbling air through alcohol. Alcohol was chosen for two reasons, (1) because the quantity of ionisation obtained with it is larger than with most liquids, and so enables us to make observations on the ions some time after formation ; and (2) because alcohol is easily obtained comparatively pure. Investigations bearing on the ionisation produced by breaking up alcohol have been made by De Broglie? and Eve.’ De Broglie showed that the electri- fication produced by bubbling air through alcohol increased very rapidly with the pressure inside the orifice through which the air issued when the pressure was over 8 ems. of mercury. He also showed that the electrification was the same with pure alcohol and with alcohol diluted with water down even to a strength of thirty per cent. alcohol, When the alcohol was diluted further, the electrification decreased. Eve sprayed a number of liquids and examined the quantity of positive and negative electrification given to the air. He found that aleohol was among the most active liquids, and that with alcohol the positive and negative electrifications were equal. ' Proc. Roy. Irish Acad., vol. xxxin, See. A (1916). 2 Le Radium, August, 1907. 8 Phil. Mag., vol. xiv (1907). R.I,A, PROC., VOL, XXXIV, SECT. A. [9] 52 Proceedings of the Royal Irish Academy. The experiments described in this paper were performed with the same method and the same style of apparatus as were used in the work on mercury. Air was forced by a pump through a fine nozzle, placed a few ems. below the surface of the alcohol. Before passing through the nozzle, the air was forced through a cotton-wool plug. In none of the experiments was the air in any way dried. A portion of the ionised air was then drawn off by means of a gasometer through the cylindrical tube designed to measure mobilities. The mobilities were determined by reading the current to the inner insulated terminal for various voltages on the outer tube, and thus plotting current- voltage curves. These curves were found to be similar to the curves obtained in the previous work; they were made up of a number of straight lines, showing that different types of ions were present. The mobilities of the ditferent kinds of ions were calculated from the formula b _ Glog - 27 V1 Uu where @ is the volume of air passing through the tube per second, } and a the radii of the tube and inner terminal, and 7 the length of the terminal. V, the saturation voltage of a group of ions, is determined by the intersection of two straight lines on the curve. Examples of these curves were given in the last paper, so it is not thought necessary to give any examples of the present curves. The positive current is always practically the same as the negative under the same conditions. In all the types of ions given below (with one doubtful exception) the values of the mobilities for the positive and negative ions are the same. Accordingly, no distinction of sign has been made; about’ half the observations refer to positive and half to negative ions. Preliminary Observations. The results of preliminary experiments on the determination of mobilities are given in Table I. M‘Cietianp and Noran—Zons produced by Bubbling Air. 53 TABLE I. Mobilitres in ems. per sec. wnder 1 volt per em. “017 0083 “0042 0024 ‘014 “0075 | “0044 0027 “012? | “0078 | 0045 “0019 “0172 “0078 0033 “0024 ? 023 ? 0077 -0036 0025 ? -0080 | 0043 “0069 “0035 -0087 | 0042 ? “0085 ? | “0042 2 “0081 ? 0053 ? 0087 ? 0039 ? -0069 ? -0035 ? “0071 ? Means, ‘015 0078 0040 0023 All the values obtained are given. Those obtained from curves in which there was some doubt about the exact point of bending have been marked doubtful, and are not included in taking the mean. The ions can be divided into four definite classes. The justification for placing numbers like -0033 (lowest in column 3) and ‘0027 (highest in column 4) in different classes is that we get ‘0033 and :0019 on the Same curve, and ‘0042 and :0027 on the same curve. This division is supported by much better numbers given further on. Most of these observations were taken before the importance of the pressure at which the bubbling takes place was fully recognized. The pressure is variable for these observations, and is not known with certainty. It is probably between 17 cms. and 21 cms. of mercury for most of them. It was found that the ion of mobility ‘015 cm./sec. was present only when the pressure was reduced to about 17 cms. When the pressure was increased, the fastest ion that could be detected was that of mobility ‘0078 cm./sec. A further increase in the pressure of bubbling caused the disappearance of this ion. Experiments were, accordingly, conducted on the effect of pressure. Variation with Pressure. A mercury manometer was placed between the cotton-wool plug by which the air was filtered and the bubbling nozzle so as to give the excess pressure inside the nozzle over the atmospheric pressure. The values given in [9*] D4 Proceedings of the Royal Irish Academy. Table Ila were obtained working at a pressure of 7 cms. of mercury, and those in Table II with a pressure of 11 cms. TABLE Ila. 049 020 0074 056 020 0076 050 014 0067 ? 055 ol4? 0071 0402 015? Means, “052 “018 “0074 TABLE ITd. “043 “016 0072 “0032 “053 “O18 ] “0071 “0040 043 ? 015? “0037 Means, ‘048 017 0072 “0036 A few curves with a pressure of 4 cms. of mercury were taken, and the presence of an ion of mobility = ‘12 ¢m./sec. was observed. From these results it is seen that the division into types of ions is further justified, and that change of pressure does not change the mobility of any of these ions. Increasing the pressure causes the appearance of types of ions of lower mobility and the disappearance of the faster ions. The total ionisation, as de Broglie has found, increases as the pressure increases. From these measurements of mobilities at different pressures, it appears that this increased electrification is due to the increased production of slower ions. The disappearance of the faster ions may be only apparent. They may be present at the higher pressures in much the same quantity as at the lower pressure, but the large increase in the total quantity of ionisation may conceal them. It is difficult to decide on this point. With this method of measuring mobilities, especially when there are so many classes of ions, it is difficult to determine accurately the fraction of any particular class present. The fraction of the ionisation due to any particular ion is got by producing the two straight lines on the current-voltage curve, whose intersection determines the saturation voltage of the ion, to meet the current axis. The length intercepted by these lines on the current axis is the current due to the M‘Cuvitanp ann Novan—Jons produced by Bubbling Air. 99 ion. Variations in the numbers, which will not seriously affect the bending point, often change slightly the slopes of the lines. Thus curves which give the same mobility will sometimes give a different fraction of the ionisation as due to the particular ion, although the total quantity of ionisation is the same. This error is all the more serious if the straight lines are short. These straight lines are short, comparatively speaking, in the present work because there are so many classes of ions, and the saturation voltages are not very widely spaced. Eyfect of Time. In the case of the ions derived from air bubbled through mereury, it was found that the mobilities observed depended on the time-interval between the bubbling and the measurement of the mobility. We have examined this effect fully in the present paper in the case of the ions in air that has bubbled through alcohol. The time-interval can be varied in two ways: by interposing lengths of tubing between the alcohol and the measuring tube, or by varying the rate of flow of the air into the gasometer from the alcohol vessel. Both methods were used, and measurements were made with a number of widely different time-intervals. The tables numbered Table III (@), (4), (ce), (d), (e), give all the numbers observed in the series of experiments performed with the object of investigating the time-effect. The pressure of air was generally between 7:4 and 9°6 cms. of mercury. On a few occasions it was as high as 11 cms. Tas_eE III. (a) Time—43 secs. — _— _ _— “0013 | 016? | 0066 | -0041 0024 | 0015 | | sGrG Ses dOSUh ipa Q; — Cj = — : Kay = 1; : 1 Kar - pin Hence, for the first part of W;", + r; Kyrie - pi- Ky = 14 VS = ————— ( ang = in mg, + ee, | 2 (ui — Mia) iH m Kiel Mid, at Mig iy Ka; — Ni It is convenient to write Kj = Ka; - Hi-p Ki! = Ka; = pi, and then the complete expression of W;* becomes vy Ge Mey) K/ Ki’ { (un; K; — MaMa K;’) os (142; Ky th iq Ki’) 5 Wii = In the same circumstances W;'* is slightly simplified, and becomes ees ay Pe et EU (ae re OC asp W; = [elm - 2) jm +7 aoe | | MESA a .. (4) | for the necessary correction is i re .9 of M Mu a 2 iF e Whe eS ( + 2) | 873 8 (a; = mri and this is zero by (3). 3. Instead of the conjugate planes through C;_, (¢,, 0,0) and C;(c, 0, 0) let another pair be taken through Cj_,(¢/, 0,0) and C/(c;/, 0,0). If, for the same ray, U; correspond to W,, in the same way v; $$$ { (am Df = py Di’) + (uni? - equ Li}, 2 (uy — pi) Di Li { (nimi MiMi Li’) + (uy nim Li’) } Uji = PLummer—On the Symmetrical Optical Instrument, 71 where ewe , hGT; AG a of Mili Mil t 77> i i ta = L; i i L; : Li TMi L; ce tha (L; —K ) , [ae (L; = K,) (5 = =. SS = - 5) t i KOT > t-1 iy CTI ) Let the ray intersect the planes through (;,, C;, C/,, C/ in the points Cia Yer 2a), (Cy Ys %), (Cia Yip 8/4), (Cv yi, 2’). By (1), to the lowest order (W = W; or U;*), 7 Lalla , -? Ke Ts (MyM ~Ky” — pum ds ;) L - KEY. = Kia i = a7 Pi-y Ai, (6) X17 lit U4 Mj ‘i a mL ie ae = ip Wi a= AG iy M ) Mi Pian } with similar equations in (2, 7), (2, n); since, when /V;, U;® are neglected mt OW, aw hai = am; 5) Minli-y = Omi By the equation of the ray in the two media a , Pale ~ Ck G3 Ye ahi a Mt > l; mM; Nj , , s ee Eh 2 OS eat re Vey Mi Nis The neglect of the third order in m, m4,... by putting /;=/,, =1 will affect MW; in the sixth order only. Therefore to this order ie = is al Ly = Ys) KVLS (Ya = Yr) eo pie Lg =)” Ol ea Dg = Bs) ae and the (m,y) part of IW" becomes vf ; I (ea Vs (us = meg) 78 (Zi = 1G)? {Zi (yi — ui) — Li (yi - Yia)\*: [12*) 72 Proceedings of the Royal Irish Academy. - This expression is to be transformed in the light of the first approxima- tion (6). Now, (Di? (yi - ui) - Li (yi - 4)P = (Ziyi — D'y/34) - 2 (Li ye - Liye 4) (Liu; — Ligne) + (Li: — Livinl = ( Le’ yg © Lyi .F a 2 ( | = Ziyi.) ( Key; = Ek; ry) isk / es + (Ky, — Key V LiL | RWG! + 2 (Dey -— Cfye.) Lin (1 — KG? LEK) +> by y? (l- Li K") Ky LD; ")+ : DL; Yi “a- = KD’ /D; KZ"). But the first approximation shows that (Lys — Ley? ,.), G’y: - Ky) are of the third order; and KEL! — Ly Ki = (Kor - nig) (Ler — as) — (irs - wig) (Kiri -— 1) ae (uy oa Hi-y) (LZ; = Kj). Hence, the sixth order being neglected, cy Ny ky” Liye Liye, 2H, W i = —_ = = art K) Ke Re ge OE Te) L232 Lz}, 2244 woe is 7 Ke A ae a FP (Z; y= Lj 3 -) othe (8) 4. On the other hand, the transformation (7) applied to (1) gives for the variation of m, y, j a 1 fas 4 wpe , OW = er Ky Ki Li ge sB ia - Yea) ~ Ki Li yd (yi - 9) Jee Ki KY’ (Ley? = Liy2, 24-4 Oe Re | = 2r2( j = K;) Ky Zz Kr ia Vind (Zy’y, -L; K% =u 1 ” , , , “a4 : ” 4 ” , + el (Dy? — Ly.) KiSy. - ’y; - Ky) Li 8y?}. Hence by (8), when Vis developed to the fourth order only, ri (D,- Ky ewe = (Diy! - Lzy?_.) Kieu - U2 — Key) Li Sy 4 (72a es Ljzi) Kix, - (Ky; > Kix...) Li &2/, and therefore we er re(L;- K) owe r r?(L; -— K; owe Ly — Diy 4 ye,” Kin; - Ky, = meee iano re i fi-4 i ‘ Ww poy ri (L;- Kj) owe ye ap 7(L;-K;) oWe L; a ~ Lz, = K ae , Kj’, -— Kiz,, =-* Se ) az, i “1 PLumMer— On the Symmetrical Optical Instrument 73 It is now necessary to calculate WW; according to (4), and here it is clearly possible to use the lowest approximation (6) so as to express the result in terms of y;,, y/, 24,2. Thus (7) give Le’ KY Bi — k f B ge es ee) ere mini (LZ; — K;) and therefore, by (2), Ky (Lyi - Ly.) fete (Li = Ki) 7 Ieee Ce KE ST) = Mele (t= Di) r% é (Zi — Ki) (Hi — bia) Y = (Lyi - Kiyi4) | 7 Ui — Ki). Let Ri = LE" (yl + a"), Rig = KE Yi + 424), Te = LEYEE (yyy + 2/223). Then ¥?+ 7 = (fy + Re, — 27))/ 75 (L; = GY me + Ne = (K,/?R, a5 LOR ic 2h LT) | nzré (LZ; = K;)? Me4 + Ne, = me OR! + Li R — 267 L/T;) | we rit Ti — Ky. Hence m,; 3F ne SF @zZ +Z*) / (a; = Cir; BP KUARE ae Re aR Ge Tits KG (Rage = 2T;) pit’ (LE; = 1K)? is rK KR - 2reliKe'Ts + 7B, el: 7%; — 2u;L; + fils) a fo (Ke): and finally by (+) 8(L; - Ki) Wi" = apap Ki KiRf — EKy'T, +. Re, Ler — nel + Ko? ee GRR = 21K; AY be + BR, (L2 aS Zul; + iB), 1G 21% K; — B(us — mig) te" (Ri - 27; + Re)? =- 2754 (A;R2, + BR? 3r 4C,T 2 ar 2D; RRs - 4ER.,T, where - 4FR/7;) Apa y (Mite = mea) Libre = Quel + wiky)® | Lens ~ Ques + weaKs) | me { rs mer” Bey hy j By = § {05 (ui — wey) 18° - K2 (KY | wer; - Ki /ui273)} Cr= 3 (0: (ui -— eg)? — LP (KY! | peri Rs Ky / Me 7%) 74 Proceedings of the Royal Irish Academy. ee k (5: ( \Ha — Mi- 1) K; (Ler; = ze + Bi Ki) , K; (Ler; = Jy; 1; ata HiK;) | | re Met; Beg? t \ ne k \5; (4G - Mie rib L; (£27; i Qu L; aE mK) hi (L2 05 Shah + MK; i) | re mer; Mears ) i iy os Fy = 3 (0; (us - mig) 7° — ND; (KE lmer; - Ki]? }- 5, For the sake of clearness, even at the expense of brevity, as little change of notation as possible has been admitted up to this point. 1t is now necessary to make a transformation which will allow two sections of the characteristic function to be fitted together. Let Ki’y: = Ho, Kyi = Hin, Li’ yi’ = Hin’, Liyi = Hin’ with the same forms obtained by substituting (2, 2) for (y, n). The succeeding refraction can be treated in exactly the same way, by changing the subscript iintoi+1. The terms of the second order in W;,, are removed and W;,,'° is obtained by substitution. But consistently with the above transformation Kinys= Hin Lin’ = Hin, and this requires that Hy,/H; = Ki,/Ki’, Hin|Hi = Li',/L,’. Now (5) give mar? (DL; - Ky)/KiO LS = grin? (Lig, — Kigq)/ Ki Li op since the second pair of planes for i become the first pair fori +1. Also if HH, = r? (L; - Kj), which is clearly consistent with what precedes, (9) become : € awe owe | ie Re a Ni Ny = oni a ihe (10) x 3 iv ra] We 2 Ya o a G ~ Si- — Oli , oi Cry = Of; Let pi = ni? + G7 = RH, Pi- = ney" + ois” = h;,/ H?, T= Mam + Cor = T/A, and also HH; = hy;(L; - Kj), H; = hjr,;(L,;- Kj). Thus hy = HWY | r2é (LD; - Kp* = (Li - Kp hy, Kin nmi - Ki) _ tig LD h; K;” : Tin, Lag = Kin) Ti . Lin, L in 7 vr; (Z; — Ki) K; :) = aeisy Ki’ hf L” ; Tiny (Lig - Ki = Kis) % j Kin PiumMer— On the Symmetrical Optical Instrument. 75 define the suecessive formation of h;, h/; and = 4We = 4 dispi.? t Bhp? + AO N27? + 2DheZhi?p-.p., = AB Sh; pisri = AP LAO pi 7; where the coefficients 4;,..., #; have the meanings found in § 4. 6. It follows at once from (10) that ~= My =m =--., c06 Shey SHG Sc Jo= tea SOR Sabk, Po oe Steps GL Sao when the third order is neglected. Therefore to the third order, by addition, 5 4) wee A] Wie Ue > Uh = we Ue >= uh => ly I On, lj 0 On 5) é , ys owe Zz y a Wwe i aiGi =) aca (= Soi =) casas J c [e J 0g; where 7 is the number of the final refracting surface, _ Ayre = - 45 We t=1 Il Ap? + Boj? + 407,77 | 2Dpyp; - 4 Epwtyi — 4F p77; Toj = nny, a CoGy i Po = Die F oe py = na + Be and in terms of the quantities already found (§ 4) A=ZAN}, B=SBAM, C= FC heh, D= DAZ? LH= SEZ, F = SFhjh;'. Let d;=4.,-4; be the distance (from left to right) between the surfaces 7 and 7+ 1. Then d; 7 (Gin im a) oc (a; = 7) = (@iy = e;’) 7 (a; = Cy), and, to repeat the notation, vali U0 Eas alee ee dapriacl\) oeaas A) Tree Coy ) a} 1 1 tail! 1 Rae oem aie oe f = - ¢ 2 us Bey \a; = C4 ss =) Ki = Ki; — Bip I Cle = Ky; — Mis Df = Lre- pig, Li = Lai - py 76 Proceedings of the Royal Irish Academy. ” , ads his, =, Pity L; h; ci Tixy K; a eS oe rs fa h; Tj L; +1 h; ie K; +1 hghs = (1; - 1G)4, Fi JH; = KiB, Hy (Hi = LiJL, HH = r2(L; - I % Ky; = Hm, Ky = Ani, Lif/yi = Hin’, Diy, = Hin; = For the beginning it is possible to write Yo=n H/=Ly, H,=7?(L, - K/L, h, =1,/L,,° te, = L,{/r, (4, — I): These formulae completely determine the Gaussian approximation’ NM =--- = =--- = NM =-+- = =---.= nis which can be used to find the terms of the third order. The two systems of conjugate planes are quite arbitrary. With Schwarzschild the first @, is identified with the object plane, the second CG,’ with the entrance pupil. The last plane C; of the first set is then the image plane in the Gaussian sense, and the last C; of the second set coincides with the exit pupil. Thus the limits of »,” or n; define the limits of the effective pencil which forms the image. Schwarzschild remarks on the analogy of the method with the theory of planetary perturbations. But a more special analogy in the variation of constants will be seen in the’ method of Delaunay’s Lunar Theory. The essential point consists in the treatment of the second-order terms in the characteristic function, whereby the principal part is removed and only residual terms of a higher (sixth) order are left. Thus the analogy is very close. 7. The formule of § 4 are easily adapted toa reflecting system. It is only necessary to write | i, m= - | Ky = Ky; =~ 1. Ki’ = Ky; +1 Lj =La;,-1, Lj’ =D; +1. Then of ea b (2h; - Kj? + (2K? + L? - AKL) Ler? ae 1, (Kfr? - 1) ; B; = b/r2 + K2/r;, C; = br? + L2e/ri, D, = bir? + Ky(2L; - Ki)/re E; = bjr2 + D; (20; - K)/ri. F,; = br? + K,Difr;. PLumMER—On the Symmetrical Optical Instrument. (iC! Here the sign has been changed throughout, the result of reversing the sign of 7; This has been done because the reflecting surfaces are usually concave to the incident ray. With this change of sign = 1 iil 1 Ka ———$ —- —~— = ~ = —— i: Gay te =e or, if A; is the vertex of the mirror and (;,, C; the points where the conjugate planes meet the axis, Ge WC Ab = ie = Wh S Uh An After the reflexion it is perhaps most convenient to reverse the axis of w at -the vertex 7; Let d; be the (positive) distance between the mirrors 7 and a+1. Then, ¢; ¢;/ having double meanings for the two directions of the axis, d; i] C Ain, + C; A; = (ea = ¢;) + (a; = e;) Cy Aj + OYA; = (Giz, = 67) + (G4; = 7). Il The necessary changes in the remaining formulae give BU SIG Ry, WOO = Ga ! is.) Il & a = | is | S a , “ hing ie Jb = I Wi te GT “ To —>————————— h; T Lint tl he fis Ukuc(Psy te ll h Unt Ti Lb. + | ape il = ———., hf = — 1 1+—., hihi = -—s pencnirkes Wine (CE =e) Se Sate / Fig Kenting tl Hii, — Lanting + 1 HH; Kr; =i” H, < La; -1 _ = REG, = 16) ah ; ae L ie = a iil = = (Lyn 1) aH (le =), No further alteration is required. Schwarzschild has given a treatment of the mirror system which is independent of the previous theory for a refracting system. R,J.A, PROC., VOL. XXXIV, SECT. A. {13} 78 Proceedings of the Royal Trish Academy. 8. The theory of the errors of the third order is contained in the equations a) Wie 4) Wwe UA 6S ee GG == ag’ -4W* = Ap? + Boj? + 407.7? + 2Dpop;’ - 4£por,; - 4Fo/'r,;, and their effects in the image plane follow immediately. But it is convenient to consider the incidence of the ray on a parallel plane at a small distance e before the image plane. The displacement of the point of incidence is given by AY) cy En Bins DB) ta € Gq -G \ LY Ky") pa? (L, - 15)’ Anj = -€. Gi ey ae eee ur Lj — KG) H; 1y77°(L; — Kj) The second term depends only on the object-point. It represents a linear distortion, or, in other words, a simple change of scale value. Therefore it needs no further consideration. Let the ray now meet a sphere of radius 0, such that the image plane is the tangent plane at the point where it meets the axis. Then 2pe = y,* when z = 0, and the first term becomes il A Ky Puy Se" HH njenj Zur (Lj — Kj) Hyp 2njr7? (L; — Kj)p =- Noni 2p. The assumption that z,=0, and therefore z;=0 to the first order, does not restrict the generality. Similarly AG; = - °Si'/2nyp. Now the errors which have the same form are those multiplied by C and J). When Z, = 0 they are nj - Mp = (20 + D) norng G = DD ne Gs If p=p, where ; I/p, = 2uj(20 + D) the image clearly becomes a straight line transversal to the axis, for nj — mo + An; = 0, PLummMrer—On the Symmetrical Optical Instrument. 79 This gives the locus of the primary focus. If p = p, where Vp, = 24D the image becomes a straight line radial to the axis, for G - % + AG =0. This gives the locus of the secondary focus. Finally, if Qu; (20 + D) - p++ 2un;D - pt =0 p> = 2uj(C + D) = 4 (py + po") it follows that ny — Mm + Any = Cng’nj’, oi — Sy + AC; = - CnrGy. This gives the spherical locus of the circles of least confusion. If C = 0 astigmatism disappears, but curvature of the field remains. The quantities C and D are the pair most easily expressed. For, when b; is omitted, 2 zt Zi 2 NB ( ki Ki, ) 2(L; - K;)?7; My” me Lhe (ui = i) / as: t= Bin) roa (its ripe Noe praia Que? (L; - K)?r; Aare (i + pea) ~ mites), and = =Dh Ph ge sy K; ; (Ler; = 2H Li + MK; Ler; = Qu Li; + mi lt;) = 75, = Ky : [ese me a I; (mi = mi) 172, 2 = TREO = Kj)n; Lier (Mi + Mig) = Miptig (20; - K)} Ai = Me SG) piss ee 2 Mili? i Hence the meaning of the Petzval condition is expressed by 1 1 3 0=32( - —| =2(0-D)=—- (= Z =). Ti \Mi Mi In the absence of astigmatism C=0, p,=p,=p, and this condition secures flatness of field. But the independent condition for flatness of the best possible field is Sale aan 2(L, Kr MLE; (Mit mig) — dein (LP + 2.0L; - 1). tMi-y i vry [18*] 80 Proceedings of the Royal Irish Academy. 9. The theory has been built up by considering the successive refractions and their combined effect. For some purposes the intermediate steps are unnecessary, and a general qualitative theory of the symmetrical instrument can be based on the consideration of the first and final ray alone. The equation (1), now written in the form OW = — py (yjdM; + 2O2j) + pay (YodIM + % 82), is general. The equations, corresponding to (7), ry = ec) wry (Lj — Kj) Hor '(L, ~ K) _ Ky Lj" (xf - 3) _ K/L, (30 - %) = — n, = 14 1 pyrP (Ly - IG) por? (L, - K,)’ represent a general transformation. Here (¥', 2%’), (y,,%) are not strictly the intersections of the ray with the planes of the entrance and exit pupils, but coincide with them to the first (Gaussian) order. The result for the variation of m, y is, in the notation of § 6, Ki Lj / KYL) - > - a >) 4+ Say as ee oe } ) 1} (Lj = kK) a(yj %) ig O25 z i) (ye L “Hi, L ak pe ae 18 GPS ales os YE) njonj + noon + 13( Kj’ B; "j a * No ) oW = i} - njeony = NON + ony, + ny Ono Mong Eh ga Rall 1 a ie A 5 — Mo(Mj - nsf: Therefore, if S = W+ no(nj - no) + So(Sy - 50) Se ( LH, 2 2\ _ LH, 2 2 2 | KH; (n; x oy ) Wf a Me No + Gi) ; the variation of S depends on (,, Z,), (nj, Z/) only, and os os 1G = 6) any” Gy = Ls =- aC : ra (11) UNG OSC oe Now S can contain no terms of the second order, because ny=%,... to the first order, And, owing to the symmetry of the instrument, S can be developed in terms of the three axial invariants pp, pj,and 7. The meaning of certain classes of terms in S is now easily seen. PLumMEerR—On the Symmetrical Optical Instrument. 81 10. Thus the terms of the form 3 4’p,2 are independent of (n,/’, 2). They disappear on differentiation, and have no effect on the image points. Similarly, terms of the form SB’p;“ are independent of the position of the object point. They constitute the whole of the spherical aberration, ‘Terms of the form 3H’p,¢r,; ave linear in (y/, Z). On differentiation they give rise to a pure distortion which affects the position, but does not disturb the quality of the point image. The remaining terms, which are represented among the third-order errors, and which have not been specially considered, are of the form S2"p;%r,;. When ¢,=0 they give i — Ny =~ NAL" pjft (1 + 2qnj?/p;') G == 2 of. 2on; Glos. These terms represent coma. The corresponding image formed by any zone of the instrument is the twice-traced circle [ny - np {1 - SF’ pj4 (1 + 9)} P + 6? = (Sq h' p42 )Png”. In the third order qg=1, and this circle is {mj — my (L = 2Foj')}? + 6? = F?no’p;”, and the whole system touches two straight lines meeting at 60° in the first-order image. This is the characteristic error of the single parabolic mirror for an object at infinity, and might be called the parabolic, as distinguished from the spherical, aberration. If there is no spherical aberration, S contains no terms independent of (m So), and therefore oS a when »=&=0, wn =¢;=0. These points on the axis are stigmatic points. = (0), If there is no coma, S contains no terms linear in (yp, Z)). Then as CHS : GAS) GS) OnjfOn Onjol, OG/dn, O6/0%) and therefore the point (mn) + 8%, ¢)+96€)) to the first order in dy, 02%) is represented stigmatically if 1») = =0 is so represented. In this case the points »,=2,=0, n;=%;=0 are aplanatic. The second condition alone can be written Ong. Gay Oy . Gax TENG Ge ty 9 Bae Tie (12) 82 Proceedings of the Royal Irish Academy. for all values of (»,, ,). But when m=% === 0, = %=H=H=0, and Dp SUCH hy iy ee mg e Yo K/L EGE tie Wong ey Ge wees uy GF py Kj" Ay ij Cea SR Hence if mM; nN fo Ky HF, tee > eka ae (13) it follows that =, 6 =%, and the conditions (12) are clearly satisfied. Now Kj" Hy _ We Yo _ Yo _ % Khan we et] to the first order, and this ratio is the reciprocal of the magnification. Hence (13) represents the sine condition, and when it is satisfied the instrument is free from coma or the effect (of any order) which depends on the first power of the coordinates of the object in the plane perpendicular to the axis. 11. The equations (11) ave general and reduce the problem of calculating the errors for any order to the development of the function S. For a single refraction omitting the similar terms in 2, Z, KK K;" (Liv? Ye Ley? 2yi- 1 , , | Refit yea ey BE Lt) \ Ir? (L; - Kp) Kk; K;’ HCG iy vy PU pee Pes 3 Li; | Tay ~ oH? "+ Fer ~ Mal Fe + Ht (ni — 144). S;= W; - Hin Lin Hi Di’ eek Gg a Hy eg Therefore in the sum, since W=3W,, H; HL,’ Vp) -F Re (no? + &7) 1S,=W-} vi : (ni (me - Wey) + Gy (Ge - S2)} = § + SS {(nieg — moni — 19-4) + (Or, - So) (6 - i-)}- =1 PLumMMER—On the Symmetrical Optical Instrument. 83 Hence as; (3 OS se 2 as; Seat. J S = ’ ra) Una m1 On’ ete i=1 If the terms found in § 3 are restored to WV;", omitting the corresponding terms in 2, S, = W,- We + (Lig! - Lig’ 4) + (Ey: - Keying)? LL ET’ = Lye - Lyi 4) (BE yi - KY 1S Lie |B KEEGY] = pra) HP? = W, — Wi IG rs \(ne as) & (Gi — Se He = ti) HP + DEL ir { (ne = i)? + Ge = Se4)") /2 (ue = pia) He? — Ky Liry{ (ng = ne) (ni = ni) + Gi - 24) (Si - 2/4) }/ (a; — wea) H, Hy Thus S¢=0, S*=JV", and for the sixth order, giving the errors of the fifth order, etd , OUP @ al We 8 (ia *) St =>] Wri + Wii eS ast : | "Ong | On’ (z, ) | nace ce - ‘ t=1 (/@ W; iw ow. 2 KiKi"; LE WWD (me rey) TER ‘ a le MH (He ~ mex) owe awe \2| | : as Mine => Nea) JERE ar L; L; v% ( any ya ( a; yy (i in) t jaw OW A) We 0 We | ln 7 ) HH; | 7 Mya 423 |, = KL’ ‘; £ = : “VG, Om Ga Oa i’ since it is possible here to use the first approximation y y Go =... — Cy = = a mes = AS; eee ni Gan Ui soe Gj» yY PS A ge est aa, of ae Sd oer In calculating W;’' the expressions (2) will no longer suffice to represent the law of refraction. To the third order the expression \( +5 Ale Zl-n & 2), ot 8 ye. ze) | | | ede ae a sR Y?+ 72 alee + 2) m2 + n2 \ + Gaal Be) isaminimum. Therefore VW VAR Ai). \ // 0 = wm; — wi. M:-, + ala = be) oy res (1 ats b) (ns = Hr) 4 ( a 2) ( : ‘) ae uy | MU; NM | = Mi | MA1- + 74> 5 27; K ; Mi4 1 1) with a similar equation obtained by interchanging (Y,m) and (Z,x). The 84 Proceedings of the Royal Irish Academy. first approximation can be used in the terms of the third order. Hence Va Milly = Mig Mey € * M) ry Mi Mia Z _ _ Milli = Mig Nig (1 4 it) re Hi — Mey ; where 1+ b; 9 M = =~ { (ui = ey)? + (mis — pe-q:-)?} 2 (14; — iq) — {ui (mF + PF) — wig (Mg? + M-4*)} = Sunes (as — meal? (Cm — M44)? + (M4 - 4)*} + AD; (mag — pig)? {Quam — pig)? + (Ha: - pigMia)?*}- The last line vanishes with 0; for spherical surfaces. The errors in the point of incidence (Y,Z) are now of the fifth order, and will only affect the calculation of W; and the optical errors of the ninth order. These indications will suffice to show that the calculation of the nine errors of the fifth order for a symmetrical optical system, though complicated, would present no serious difficulty. That the caleulation has not apparently been made doubtless results from the practical difficulty of reconciling the removal of the errors of the third order with the conditions of achromatism —a difficulty which hitherto has not been completely overcome. 12. It may be convenient to add a brief summary of the contents of this paper :— § 1. The modified characteristic function. 2. Refraction at one surface. Form of the second order terms W;". 3. Introduction of new variables, Transformation of. IJ’;". § 4. Elimination of second-order terms. Calculation of #V7;,'". . Changes in the coordinates. a Fist sao Extension to any number of refractions. Formule adapted to a mirror system. Astigmatism and curvature of the field. The Petzval condition. The theory in most general terms. ae LP ma — SO # . Spherical aberration and coma, The sine condition. § 11. Further development. Preparation of formule for calculating errors of the fifth order, PROCEEDINGS OF THE ROYAL IRISH ACADEMY VOLUME XXXIV SECTION B.—BIOLOGICAL, GEOLOGICAL, AND CHEMICAL SCIENCE, DUBEIN: HODGES, FIGGISS& CO.) Erp. LONDON: WILLIAMS & NORGATE 1917-1919 THE ACADEMY desires it to be understood that they are not answerable for any opinion, representation of facts, or train of reasoning that may appear in any of the following Papers. The Authors of the several Essays are alone responsible for tveir contents. Dustin : Pumten at toe Ustvensrry Puess ny Ponsonpy anp Gress. CONTENTS SECTION B.—BIOLOGICAL, GHOLOGICAL, AND CHEMICAL SCIENCE. PAGE Atear (Joseph), D.So. :— See wnder Ryan (H.). Carpenter (G. H.), D.Se., and F. J. 8. Pottarp :— The Presence of Lateral Spiracles in the Larva of Hypoderma. (Plates VIII-XTII1), 6 5 : : : : 6 5 WR Drumm (J. J.), M.So. :— See wnder Ryan (H.). Guover (T.), M.Sc. :— See wnder Ryan (H.). Newton (Hi. T.), F.R.S. :— See under Souanrrr (R. F.). O’Connztt (P.), M.Sc. :— See under Ryan (H.). O’Rrorpan (W. M.), M.Sc. :— See wnder Ryan (H.). Poxtarp (F. J. 8.) :— See wnder Carpenter (G. H.). Porter (Lilian), M.Sc. :— On the Attachment Organs of the common Corticolous Ramalinae. (Plates II-IV), ¢ : : > ° : : : 5 Alt On the Attachment Organs of some common Parmeliae. (Plates XXL-XXITI), : . : . . : . . . 205 Ryan (H.), D.Sc., J. Angar, D.Sc., and P. O’Coynett, M.Sc. :— Syntheses of some new substantive Dyes derived from Benzidine- Sulphone, 5 : 5 - : : : : - > Ble Ryan (H.), D.So., and J. J. Drumm, M.Sc. :— The Nitro-derivatives of Phenyl-2-Naphthylamine, . . : = 265 Contents. PAGE Ryan (H.), and T. Grover, M.Sc. :— On the Nitro-derivatives of Diphenylamine, . : : , ae); Ryan (H.), and W. M. O’Riorpan, M.Sc. :— Ona, 8, and y Trinitrotoluenes, : . a AllT/3 On the action of Bromine on some ckeatiess of Diphenylamine . 218 Ryan (H.), and Phyllis Ryay, M.Sc. :— On the Action of Nitrous Acid and Nitric Acid on Diphenylamine. I, 194 ” 3 9 % IW Pal Ryan (Phyllis), M.Sc. :— - See under Ryan (H.). Sonarrr (R. F.), B.Sc., Pa.D.; H. J. Seymour, B.A., B.Sc.; and E. T. Newton, F.R.S. :-— The Exploration of Castlepook Cave, County Cork. (Plates V-VII), 88 Seymour (H. J.), B.A., B.Sc. :— See under Scuanrr (R. F.). Srepuens (Jane), B.A., B.Sc. :— Report on the Sponges collected off the Coast of Ireland by the Dredging I!.xpeditions of the Royal Irish Academy and the Royal Dublin Society. (Plate I), . : : ' : : : 1 Stepuenson (T. A.) :— On certain Actiniaria collected by the Irish Fisheries Department during the years 1899-1913. (Plates XIV-XX), . ; - 106 ERRATA. SECTION B. p. 65, 1.9 For milk pre-molar read milk molar. p- 17, last line. For 20 read 12. PROCEEDINGS THE ROYAL IRISH ACADEMY PAPERS READ BEFORE THE ACADEMY Je REPORT ON THE SPONGES COLLECTED OFF THE COAST OF IRELAND BY THE DREDGING EXPEDITIONS OF THE ROYAL IRISH ACADEMY AND THE ROYAL DUBLIN SOCIETY. By JANE STEPHENS, B.A., B.Sc. PLATE I. Read June 11. Published Juny 18, 1917. A SHORT account is given in the following paper of collections of sponges obtained a good many years ago by the Royal Irish Academy and by the Royal Dublin Society off the west and south-west coasts of Ireland. These sponges were deposited in the National Museum, Dublin, where they have since remained unnamed, with the exception of two or three, which were referred to in the report on the marine sponges of the Clare Island Survey (12), as they were the sole representatives of their species known within the area of that survey. Excepting the first cruise of the “Poreupine” in the year 1869, the earliest attempts at deep-sea dredging off the west and south-west coasts of Treland were carried out by the Royal Irish Academy, which equipped three dredging expeditions in the years 1885, 1886, and 1888. Accounts of the invertebrates then obtained, with the exception of the sponges, were published in the Proceedings of the Academy (4 and 5). Later on, in the years 1890 and 1891, the Royal Dublin Society organized a survey of the fishing grounds off the west coast of Ireland. In the course of this survey large collections of invertebrates were made. A preliminary list of the species was published in the report on the Survey (8), in which mention was made of about half a dozen of the commonest sponges. R.I,A. PROG., VOL. XXXIV,, SECT, B. [A] 2 Proceedings of the Royal Irish Academy. Several of the groups have since been worked out in considerable detail, but again the sponges were left untouched. The number of sponges in these united collections is not great. In fact, as always happened in the earlier dredging expeditions, the only sponges that were collected as such were the larger and more conspicuous kinds. Shells, corals, stones, and other material dredged up were not examined and retained, as they would be to-day, for small, encrusting sponges, which are now known to exist in extraordinary numbers and in great variety of species. However, a search through other groups in the Royal Irish Academy and Royal Dublin Society collections in the National Museum, such as the mollusca, the hydroids, and particularly the corals, resulted in the discovery of a number of small, for the most part encrusting, sponges of considerable interest. The number of species in all is only thirty-eight. Of these two, Wycale (Paresperella) atlantica and Forcepia fragilis, are described as new. The former is interesting on account of the fact that it is the first representative of the sub-genus Paresperella found in the Atlantic Ocean. The following five species are noted for the first time within the Irish area :— Laxosuberites ectyoninus Jopsent. Desmacidon fruticosum (Montagu). Hymenancora conjungens Lundbeck. Microciona laevis Bowerbank. Tragosia polypoides (Schmidt). Two of these species, Laxosuberites ectyoninus and Hymenancora conjungens, have not been taken previously off any part of the British Isles; each, in fact, has only been recorded once up to the present, the first-mentioned in the Mediterranean, the second off the south coast of Iceland. CALCAREA. The five calcareous sponges represented in the collection are among the commonest species found round our coast, and call for no special remark. They are as follows :— Clathrina coriacea (Montagu). R.D.S. Survey. Blacksod Bay. Leucosolenia complicata (Montagu). R,D.S. Survey. Station 238, Lough Swilly, 8-12 fathoms, SrepHens— Report on Sponges collected off the Coast of Lrelund. 3 ; Sycon ciliatum (Fabricius). R.L.A. Exp., 1886. Log 41, Crookhaven, 35 fathoms. R.D.S. Survey. Blacksod Bay; Kilkieran Bay; Clare Island; Station 25, rock pools, west shore of Killeany Bay; Station 158, Greatman’s Bay, Co. Galway; Station 238, Lough Swilly, 8-12 fathoms. Grantia compressa (Fabricius). R.D.S. Survey. Kilkieran Bay; Clare Island; Smerwick Harbour; Station 167, Blacksod Bay, 3-4 fathoms; Station 238, Lough Swilly, 8-12 fathoms. Leucandra nivea (Grant). R.D.S. Survey. Blacksod Bay, 5 fathoms. NON-CALCAREA. Order MYXOSPONGIDA. Halisarea Dujardini Johnston. R.D.S. Survey. Station 240, Lough Swilly, 6-84 fathoms. This common species is represented by one specimen growing on the leg of a crab (Hyas avaneus), which is thickly covered with Polyzoa and Hydroids. Order TETRAXONIDA. Family 'THENEIDAE. Thenea muricata Bowerbank. R.J.A. Exp., 1888. Log 69, 51° 1’N., 11° 50’ W., 750 fathoms. The only specimen in the collection is 45 mm. by 38 mm. in diameter. It has one large osculum placed almost centrally on the upper surface of the ~ sponge. A large number of specimens of this species have been obtained within recent years by the Irish Fisheries Branch off the south-west coast of Treland at depths varying between 625 and 982 fathoms (18). Family PACHASTRELLIDAE. Poecillastra compressa (Bowerbank). R.L.A. Exp., 1886. Log 53 (or ? log 20). Otf the Skelligs, 70-80 fathoms, This species is represented by one small fragment. Like Thenea muricata, numerous specimens of this species have been dredged within recent years off the south-west coast of Ireland at depths of 468-728 fathoms (13). These depths are considerably greater than that at which the present specimen was [A*] aes Proceedings of the Royal Irish Academy. obtained; but the species is known to occur in shallower water, for example off the north-west coast of France at about 28 fathoms, and off Rackall at 60 fathoms. Family GEODIDAE. Pachymatisma johnstonia (Bowerbank). R.D.S. Survey. Station 86, off Inishmaan, 20-7 fathoms. This species is doubtfully recorded by Mr. Holt (8). There is no specimen from this station in the Museum, so that the identification cannot be verified. The species is, however, well known round the Irish coast between tide-marks at extreme low water, as well as in rather shallow water-dredgings. Family CLIONIDAE. Cliona celata Grant. R.D.S. Survey. Station 32, Birturbuy Bay; Station 196, Davalaun Sound, 16-15 fathoms ; s.s. “ Harlequin,” neighbourhood of Cork Harbour. All the specimens in the collection are the massive form of this common boring sponge. The specimens from station 196 are recorded by Mr. Holt (8) as “great masses of sponge (haphyras Griffithsii),” a broken piece of which measured 20 inches by 17 inches by 8 inches. Cliona vastifica Hancock. R.I.A. Exp., 1886. Log 49, 51° 20'N., 11° 26’ W., 42 miles from Great Skellig, 160 fathoms. R.D.S. Survey. Station 124, 50 miles W. of Bolus Head, 220 fathoms. Both the specimens are boring in coral. The sponge from the first- mentioned station is boring extensively in a specimen of Caryophyllia clavus, that from the second locality is in Lophohelia prolifera. Only a very small piece of the sponge was found in the Lophohelia. Alectona Millari Carter. R.D.S. Survey. Station 124, 50 miles W. of Bolus Head, 220 fathoms. This sponge was found boring in coral (Lophohelia prolifera), Both this and the preceding species have recently been recorded for the Irish area (13), . having been obtained in the course of the dredgings carried on by the Irish Fisheries Branch. Family SUBERITIDAE. Laxosuberites incrustans Stephens. R.LA. Exp., 1886. Log 49, 51° 20'N., 11°26’ W., 42 miles from Great Skellig, 160 fathoms. Srepnens—Reporl on Sponges collected off the Coast of Ireland. 5 The single small specimen in the collection is growing on a coral (Caryo- phyllia clavus). This species was recently described (13) from specimens dredged by the Ivish Fisheries Branch off the west and south-west coasts of Ireland, all of which, with two exceptions, are also growing on Caryophiyllia clavus. Laxosuberites ectyoninus Topsent. RDS. Survey. Station 124, 50 miles W. of Bolus Head, 220 fathoms. The two specimens form sma!l thin encrustations on coral (Lophohelia prolifera). The spicules agree exactly in shape and size with those of the type, of which a detailed description and figures are given by Professor Topsent (16). The species has been obtained only once previously, namely off Banyuls, at a depth of 500-600 metres. Suberites carnosus (Johnston). R.I.A. l’xp., 1888. Log 70, Berehaven, 5 fathoms, one specimen. R.D.S. Survey. Blacksod Bay, two specimens: Station 33, Clifden Harbour, 6 fathoms, two specimens; Station 133, Dingle Bay, 40 fathoms, two specimens. This species is widely distributed round the Irish coast, but is not very common. Ficulina ficus (Linn.). R.A. Exp., 1885. Log 3, 40 miles off south-west Iveland, 51°15’ N,, 10°31’ W., 80 fathoms, one specimen. R.I.A. Exp., 1888. Log 70, Berehaven, 5 fathoms, one specimen ; log 73, probably off south of Ireland, 50 fathoms, one specimen, R.D.8. Survey. Blacksod Bay, eight specimens; station 35, Clifden Harbour, 6 fathoms, one specimen and fragments ; station 73, Killary Bay, 15-9 fathoms, three specimens ; Kenmare River, three specimens; station 114, off the Skelligs, 80 fathoms, one specimen. The usual forms assumed by this polymorphic sponge are represented in the collection, namely, encrusting, massive, and “suberea” forms. ‘lhe last- named, which covers shells inhabited by hermit crabs, is very common off the coast. Mr. Holt (8) records the species, under the name Suberites domunculus, for station 38, Blacksod Bay. Family HAPLOSCLERIDAE. Sub-family GELLIINAE. Gellius flagellifer Ridley and Dendy. R.D.S. Survey. Station 124, 50 miles W. of Bolus Head, 220 fathoms. One small specimen belonging to this species is growing on coral 6 Proceedings of the Royal Irish Academy. (Lophohelia prolifera). Gellius flagellifer has already been recorded for the Irish area (14). It was obtained by the Ivish Fisheries Branch, and was growing on [ophohelia. Gellius ravus Stephens. R.D.S. Survey. Blacksod Bay. The specimen, which is growing on an oyster shell, has been already referred to in the course of the description of the species given in the report on the sponges of the Clare Island Survey (12). It differs from the type in having shorter and more slender oxea. The species is evidently widely distributed round our shores, as it has been found at several places off the west, north-west, and east coasts. Oceanapia robusta (Bowerbank). R.D.S. Survey. Station 80, off Slyne Head, 55 fathoms. The merest fragment of a fistula of this well-characterized species is in the collection. ‘The species has been recorded previously for the Irish area (14). Sub-family REeENTERINAE. Reniera simulans (Johnston). R.D.S. Survey. Neighbourhood of Cork Harbour; west coast of Ireland. Two specimens of this common Reniera are in the collection. Reniera sy. R.D.S. Survey. Station 124, 50 miles W. of Bolus Head, 220 fathoms. A number of small specimens of a Reniera are growing on a piece of coral (Lophohelia prolifera). They are oval in outline, with a single osculum at the _ summit. They are soft and fragile in texture, and measure only 3 or 4 mm. in height. Their skeleton consists of a unispicular network of oxea cemented together at the ends by a little spongin. The oxea measure 0°18 - 0-22 mm. by 0009 mm. The sponge seems very similar in appearance and structure to Renicra spongiosa Topsent (17), but the oxea are much smaller than those of Professor Topsent’s species. Halichondria panicea (Pallas). R.A. Exp., 1886. Log 41, Crookhaven, 34 fathoms. R.D.S. Survey. Neighbourhood of Cork Harbour ; west coast of Ireland. This, the commonest sponge round our coasts, is represented in the collection by a few fragments. SrepHens—Report on Sponges collected off the Coast of Ireland. 7 Fanuly DESMACIDONIDAE. Sub-family Myca.inae. Mycale macilenta (Bowerbank). R.D.S. Survey. Blacksod Bay. ‘I'wo specimens. ‘he sponges are covering the valves of two living Pectens (P. opercularis). Mycale placoides (Carter) Lundbeck. R.LA. Exp., 1886. Log 53, 5-8 miles west of Great Skellig, 70-80 fathoms. Three small fragments. R.D.S. Survey. ? Blacksod Bay, fragments; station 72, 20 miles off Achill Head, 127 fathoms, two specimens. ‘This species is taken in the sense in which it is understood by Lundbeck (11), who carefully distinguishes Carter’s species from its near ally, Mycale lingua (Bowerbank). The largest specimen obtained measures 150 mm. by 90 mm. by 43 mm. in thickness. The specimens which are labelled Blacksod Bay have already been referred to (12, p. 34). It is probable that the locality is wrongly given on the label, as Mycale placoides has never been taken in such shallow water as obtains in Blacksod Bay, where the soundings are, for the most part, 6 to 8 fathoms, a depth of 10 fathoms occurring towards the mouth of the Bay. Since the foregoing specimens were collected, numbers of large specimens have been obtained off the west and south-west coasts of Ireland by the Irish Fisheries Branch (14). Mycale (Paresperella) atlantica, n.sp. Plate I, fig. 1. R.D.S. Survey. Station 124, 50 miles W. of Bolus Head, 220 fathoms. The sponge is growing on Lophohelia prolifera, and it formed apparently a thick encrustation or cushion on the coral. Its longer diameter is about 18 mm., but the sponge is greatly damaged, so that its exact shape and limits cannot be made out. It is well characterized, however, by its distinctive and beautiful spiculation. The texture of the sponge is soft and fragile, and the surface, seen under the lens, is very slightly hispid. A second specimen of the species, merely a small fragment, was growing on another branch of the coral. Skelefton.—As far as can be seen from the scanty material available the main skeleton is composed of fibres, up to about 0°05 mm. in thickness, which consist of closely packed subtylostyli. These fibres run upwards through the sponge, branching here and there, At a short distance below the surface they 8 Proceedings of the Royal Irish Academy. divide into finer divergent strands. The terminal spicules of these strands spread out in a fan-like manner, and their tips, at least in the preserved . sponge, pierce the dermis and project for a short distance above it. The dermal skeleton is an irregular net-work of fibres. ‘These fibres are much more slender than those of the skeleton, consisting sometimes of only two or three rows of spicules. The pores are seen in the meshes of the dermal net- work; they vary in diameter from about 0:05-0:25 mm. Spicules.—(1) The megascleres are subtylostyli; they measure 0'35- 0-45 mm. in length, and have a maximum thickness of 0:008 mm. Their shaft is slightly fusiform, and usually slightly crooked. The rounded end is sometimes fairly well marked off by a constriction beneath it; the other end tapers to a short point. (2) Anisochelae of two sizes are present ; the larger form rosettes, which are present in great numbers both in the dermal membrane and throughout the sponge. The smaller anisochelae are also scattered both in the dermis and in the choanosome. The first-mentioned anisochelae have a rather egg- shaped larger end, as the lower angles of the alae are rounded off, the lower margin of the tooth being rounded off in the same way. The free portion of the shaft is extremely short. The smaller alae also have rounded corners. The length of the rosette anisochelae is very constant, being 0:027-0-03 mm. The breadth of the larger alae is about 0°012mm. The smaller anisochelae are similar in shape to the foregoing, but only reach a length of 0°019 mm. (3) Sigmata.—These are contort; the ends of the spicule are very slender, and sharply pointed; one is bent in abruptly, while the other has a wider curve. The outer margin of the spicule is serrated towards both ends, the character which distinguishes the species assigned to the sub-genus Paresperella. The longer axis of the sigmata is from 0°1-0°165mm., and the maximum thickness is 0005 mm. The sigmata, which are often in bundles, occur in enormous numbers through the whole sponge. (4) Toxa.—These spicules are rather scarce. They occur in small bundles in the interior of the sponge. ‘They were not seen in those portions of the dermis which were examined. ‘I'he toxa are extremely slender, and are 0-08 mm. in length; the curve in the centre of the spicule is very wide. The foregoing sponge is closely allied in spiculation to Carter’s species, Mycale (Paresperella) serratohamata (2), the type species of Paresperella Dendy (8), but differs from it in the greater length of the various kinds of spicules, as well as in the shape of the megascleres, as can be seen from a comparison of the figure (Plate I, fig. 1a, 4) with that of the subtylostylus of the older species given by Professor Dendy (3, Pl. XI, fig. 22). The toxa, too, differ in shape in the two species. Srepuens—Report on Sponges collected off the Coast of Ireland. 9 The Ivish species approaches even more closely a sponge found on the beach at Vancouver Island, and assigned by Lambe (9) to Carter’s species, but which, as Professor Dendy has already pointed out (3, p. 162), is no doubt a distinct species. The same kinds of spicules are present in this sponge as in the Irish specimen, but here again the measurements are different. The subtylostyli in the Vancouver sponge are shorter and thicker, the anisochelae longer, while the toxa, differing in shape, are only half the length of the corresponding spicules in the new species. The remaining species assigned to this sub-genus are—penicillium Lendenfeld, see Hallmann (6), macrosigma Lindgren (10), moluecensis Thiele (15), bidentata Dendy (8), repens Whitelegge (19), dichela Hentschel (7). None of these possesses toxa, and all are further marked off from the new species by differences in the form and measurements of the spicules. The new species is the first representative of the sub-genus which has been found in the Atlantic Ocean, those previously described having been taken off Ceylon, or in various parts of the Pacific Ocean. Desmacidon fruticosum (Montagu). R.D.S. Survey. Station 133, Dingle Bay, 40 fathoms. Several pieces of this sponge, now recorded for the first time within the Irish area, are in the collection. ‘The largest piece is 80 mm. in height and 105 mm. in its greatest breadth. It is very similar in shape to the specimen figured by Bowerbank (1, vol. iii, Pl. LXI). Forcepia fragilis, n. sp. Plate I, fig. 2. R.LA. Exp., 1888. Log 69, 51° 1’N., 11° 50’ W., 750 fathoms, The sponge is not attached to any support. It is somewhat oval in outline, and measures 25 mm. by 18mm. by 15mm. It is very fragile in texture, and its surface is damaged. Skeleton.—The main skeleton consists of a loose, irregular network of styli. Sometimes only two or three spicules lie side by side to form the meshes, but usually the styli are multiserially arranged. The arrangement of the dermal skeleton cannot be made out, as the surface of the sponge is rubbed away. Spicules—(1) The styli are slightly and somewhat irregularly curved. At one end they taper to a short point; the other end is rounded. They measure 0:6—-0'77 mm. in length by 0-018-0-021 mm. (2) The dermal spicules are tylota, measuring 0°4-0°45 mm. in length by 0:005-0:008 mm. One end of these spicules is sometimes rather more rounded than the other, but the difference is slight. R.I.A, PROC., VOL. XXXIV,, SECT. B, [B J 10 Proceedings of the Royal Irish Academy. (3) The isochelae arcuatae have a slightly curved shaft; they are 0:024— 0°033 mm. in length. ‘The tooth is rather narrow, and is about the same length as the alae. (4) Forcipes.—These are all of one kind, and, measured from the curve to the end of the longer leg, they are 0:058-0:045 mm. in length, and at the most are about 0002 mm. in thickness at the curve. The legs are slightly divergent ; one is a little longer than the other, and each terminates in a small button-like knob. Sometimes the spicule is twisted so that the legs cross each other. The foreipes are very minutely spined. Under a high power of the microscope it can be seen that the spines are arranged in rows, and that the points are directed towards the curve of the spicule. (5) The sigmata are plane, and have a longer axis measuring 0°13- 016 mm. in length. Their thickness is 0-006 mm. All the forms of the microscleres are present in great abundance throughout the sponge. Forcepia fragilis n. sp. may be compared with the sponge assigned by Professor Topsent (17) to Forcepia bulbosa (Carter), which Lundbeck has decided is a distinct species, and to which he has given the name Forcepia azoricea (11, Part I], p. 210). Judging from the description given by Professor Topsent, the foreipes in his sponge are a different shape and size from those in the new species, and, what is more important, more than one kind of forceps appears to be present in the specimen from the Azores. In addition, the styli of Yorcepia azorica are sometimes slightly spined, while the styli of the Irish species are quite smooth. The spiculation of Fbreepia fragilis is very similar to that of Forcepia Thielei Lundbeck (11), in which species only one kind of forceps is present, but this spicule differs in shape from, as well as being longer than, the forceps of the new species. Grayella sp. R.D.S. Station 124, 50 miles W. of Bolus Head, 220 fathoms. There are about half a dozen small encrusting specimens of a species of Grayella growing on a piece of Lophohelia prolifera. They are much con- tracted, and their pore-bearing areas appear like small papillae on the surface of the sponge. Their spicules measure as follows:—(1) Tornota, 0:25 -0:325 mm. in length; (2) Acanthostyli, 0:1-0-18 mm. in length; (3) Isochelae arcuatae, 0:019 — 0-021 mm. in length. The foregoing measurements agree very well with those given by Professor Topsent (18, p. 46) for several encrusting specimens of Grayella obtained off the coast of Belgium, the north of France, and in the Gulf of Lions, which SrepHuns—Report on Sponges collected off the Coast of Ireland. 11 specimens Professor 'Topsent is inclined to think may have to be assigned to Grayella pyrula (Carter), a species which is typically pedunculate, and which has, typically, considerably longer megasclera than these sponges possess. Professor Topsent states that he has tried to establish a distinct species for these specimens, but found there was too little difference in the size of their spicules and those of a globular Grayella which he had referred (17) to Grayella pertusa (Lopsent), and which Lundbeck later regarded as identical with Grayella pyrula (11). It may be stated here that encrusting specimens of Grayella pyrula are known. As Professor Topsent says, the species of Grayella must be increased in number to a large extent or a creat variability must be admitted in Grayella pyrula. It would be necessary to study a large series of specimens before these questions could be decided, so that the only course is to leave the Irish specimens unnamed for the present. The species Grayella pyrula (Carter) has been taken off the Ivish coast (14). Sub-family Ecryonrnag. Hymedesmia paupertas (Bowerbank). R.D.S. Survey. Station 124, 50 miles W. of Bolus Head, 220 fathoms. The sponge is growing in a small patch on coral (Lophohelia prolifera). The species was first obtained off the Irish coast in the course of the Clare Island Survey (12), and it has since been taken at other stations by the Irish Fisheries Branch. Hymedesmia pansa Bowerbank. R.D.S. Survey. Blacksod Bay. The sponge is growing in a thin enerustation on an oyster shell. A description of it has been published in the report of the sponges of the Clare Island Survey (12). Hymedesmia Dujardini (Bowerbank). R.D.S. Survey. Station 124, 50 miles W. of Bolus Head, 220 fathoms. Several small specimens are growing on Lophohelia prolifera. The species has previously been obtained on several occasions within the Ivish area. Hymenancora conjungens Lundbeck. R.D.S. Survey. Station 124, 50 miles W. of Bolus Head, 220 fathoms. The sponge is growing in two patches on Lophohelia prolifera. This species was taken only once previously, namely, to the south of Iceland in 296 fathoms (11). The spicules of the Irish specimen agree [2B] i2 Proceedings of the Royal Irish Academy. exactly with those of the type, except that the large acanthostyli are rather shorter, their maximum length being 0-31 mm. as against 0-41 mm. of the first found specimen. This is the first time that a sponge belonging to the genus Hymenancora has been found within the Irish area. Microciona laevis Bowerbank. Plate I, fig. 3. R.D.S. Survey. Station 124, 50 miles W. of Bolus Head, 220 fathoms. The sponge is growing in two small patches on pieces of coral (Lophohelia prolifera). Its surface is very hispid. As far as can be seen from the scanty material available for examination, the dermal spicules are collected together in vertical brushes, their ends projecting above the surface of the sponge. The skeletal fibres are very short and plumose, the smaller styli being placed towards the exterior of the fibre. The long styli project far above the surface of the sponge. There is a considerable quantity of spongin round the bases of the styli. Spicules—(1). The skeletal styli vary very much in size. The largest reach a length of about 1-4mm., and are 0-024 mm. in diameter above the head. The shaft is slightly curved and tapers towards one end to a rather short point, and at the other end the rounded head is slightly constricted off from the shaft, and is smooth or is set with a few minute spines. The smallest styli are about 0-14 mm. in length, with a diameter of 0007 mm. above the head. They are of the same shape as the largest styli, but are furnished with minute spines along the shaft, as well as having the head well spined. These two extremes are united by styli of intermediate sizes and degrees of spination. (2) The dermal styli have a very slender, sometimes rather crooked, shaft, and the head is microspined. Their length varies from about 0°5 to 0-78 mm., and the maximum diameter is 0°006 mm. (3) The toxa are scattered in great abundance throughout the sponge. Their shape is very constant, but they vary greatly in size, from extremely minute to about 016mm. in length, with a thickness of 0:°004mm. The curve in the middle of the spicule is even, and not very abrupt, and the ends are slightly re-curved, and sharply pointed. This species has been taken only once previously, namely, off the Shetlands (1). Plocamia microcionides (Carter). R.D.S. Survey. Station 124, 50 miles W. of Bolus Head, 220 fathoms. The specimen is spreading over several branches of a coral (Lophohelia prolifera), and it is easily picked out from among the numerous encrusting Srepuens—Leport on Sponges collected off the Coast of Ireland. 13 sponges with which it is growing by its long styli, which project for about 3mm. above the surface of the sponge. Plocamia microcionides is distin- cuished from Plocamia ambigua (Bowerbank) chiefly by its very long, perfectly smooth styli instead of acanthostyl, and by its acanthostrongyla, which have stout, blunt spines, with microspined summits, instead of the sharply pointed spines of the corresponding spicules of Plocamia ambiqua. Plocamia microcionides has been found on Lophohelia dredged by the Irish Fisheries Branch (14), and it is hoped to publish a description and figures of the species in a forthcoming paper. Family AXINELLIDAE. . Phakellia ventilabrum (Johnston). R.LA. Exp. Log 20 (or ? log 53), off the Skelligs, 70 to 80 fathoms. Eight specimens. R.LA, Exp. 1886. Log 53, 5-8 miles W. of Great Skellig, 70 to 80 fathoms. Twelve specimens. R.D.S. Survey. Station 12, off Dingle Bay, 53 fathoms, one specimen; Station 80, off Slyne Head, 55 fathoms, two specimens; Station 85, Galway Bay, 19-15 fathoms, one specimen; Station 225, Rosses Bay, 32-25 fathoms. Fourteen small specimens. The specimens obtained at Station 225 are small cup-shaped sponges, varying in diameter from 25 to 60 mm., while those dredged off the Skelligs are much larger, the largest having a diameter of 185mm. The specimens from the latter locality are fan- or cup-shaped sponges, most of them having the margin deeply cut into a number of lobes, the indentations sometimes reaching nearly to the base of the sponge. These specimens are referred to by Professor Haddon (5, p. 38). Mr. Holt (8) records the species from Station 115, off the Skelligs, 62-52 fathoms. As may be judged from the foregoing list of stations, the species is fairly common off the west and south- west coasts, at depths of about 18 to 100 fathoms. Phakellia rugosa (Bowerbank). R.D.S. Survey. Station 80, off Slyne Head, 55 fathoms. The species is represented by one small, stalked, club-shaped specimen, only 9mm. in height. Phakellia rugosa has been recorded once previously for the Irish area (14). Tragosia infundibuliformis (Johnston). RIA. Exp. Log 53 (or ? log 20), off the Skelligs, 70-80 fathoms. One specimen. R.LA. Exp, 1886. Station unknown. One specimen. 14 Proceedings of the Royal Irish Academy. R.D.S. Survey. Station 225, Rosses Bay, 32-35 fathoms. Three speci- meus. This sponge, which has been obtained fairly often in recent years off the west and south-west coasts of Ireland, is evidently not as common as Phakellia ventilabrwm, small cup-shaped specimens of which it resembles superficially. The two species can, with practice, be distinguished without microscopical examination owing to a difference in texture and in the margin of the cup, which is rather thick and rounded in Tragosia, and which is thinned to a rather sharp edge in Phakellia. . Tragosia polypoides (Schmidt). R.D.S. Survey. West coast of Ireland. The sponge is branching in one plane, and has a fan-shaped outline. It is 46 mm. in height by 51 mm. in its greatest breadth. This is the first time the species has been obtained within the Irish area. Hymeniacidon caruncula Bowerbank. R.D.S. Survey. Blacksod Bay. The only fragment of this very common species in the collection was preserved owing to its being overgrown by a gymnoblastice hydroid. Order EUCERATOSA. Family SroNGELIIDAE. Spongelia fragilis (Montagu). R.D.S. Survey. Station 240, Lough Swilly, 6-8} fathoms. A specimen of this common species is growing on the back of a crab (Hyas araneus), which is thickly covered with Polyzoa and Hydroida, and which has on one leg a growth of Halisarca Dujardini. LIST OF REFERENCES. 1. BOWERBANK, J. S—A Monograph of the British Spongiadae. Vols. 1-3. Ray Society, London, 1864, 1866, 1874. 2, Canter, J. H.—Report on Specimens dredged up from the Gulf of Manaar, and presented to the Liverpool Free Museum, by Captain W. H. Cawne Warren. Ann. Mag. Nat. History (5), v, 1880. 3. Denpy, A—Report on the Sponges. Jn Report on the Pearl Oyster Fisheries of the Gulf of Manaar. Suppl. Report, xviii. London, 1905. ~I 10. 15. 14. SrepHens—Report on Sponges collected off the Coast of Ireland. 15 . Hannon, A. C.—First Report on the Marine Fauna of the South-West of Ireland. Proc. Royal Irish Acad. (2), iv, 1886. Hannon, A. C., and Rey. W. 8. Green.—Second Report on the Marine Fauna of the South-West of Ireland. Proce, Royal Irish Acad. (8), i, 1888. HaAtimann, E. F.—A_ Llevision of the Monaxonid Sponges described as new in Lendenfeld’s “Catalogue of the Sponges in the Australian Museum.” Part iii. Proc. Linn. Soc. of New South Wales. Part ii, 1914. . Hentrscuet, E.—Tetraxonida. Il. Jn Die Fauna Siidwest-Australiens, von W. Michaelsen und R. Hartmeyer. Bd. iii, 1910-1911. . Hout, E. W. L.—Survey of the Fishing Grounds, West Coast of Ireland, 1890-91. Report on the Results of the Fishing Operations. Sci. Proc. Royal Dublin Soe. (N.S.), vii, 1892. Lampn, L. M.—Sponges from the Western Coast of North America. Proe. and Trans. Roy. Soc. Canada for the year 1894; Sect. iv, 1895, LinpGRrEN, N.G.—Beitrag zur Kentniss der Spongien fauna des Malayischen Archipels und der chinesischen Meere. Zool. Jahrbiicher (Syst.), xi, 1898. . LunpBeck, W.—Porifera. Danish Ingolf Expedition, vol. vi. Part i, 1902; Part 1, 1905; Part in, 1910. Copenhagen. 2. STEPHENS, JANE.—Marine Porifera of the Clare Island Survey. Proc, Royal Irish Acad., xxxi. Part 59, 1912. STEPHENS, JANE.—Sponges of the Coasts of Ireland. I. The Triaxonida and part of the Tetraxonida, Fisheries, Ireland, Sci. Invest., 1914. LV [LOWS i: STEPHENS, JANE.—Preliminary Notice of some Irish Sponges. The Monaxonellida (Sub-order Sigmatomonaxonellida) obtained by the Fisheries Branch, Department of Agriculture and Technical Instrue- tion, Ireland. Ann. Mag. Nat. History (8), xvii, 1916. . THIELE, J.—Kieselschwamme von Ternate, ii. Abhandl. Senckenbere- ischen naturf. Gesellschaft, xxv. 1903. . Torsent, E.— Etude Monographique des Spongiaires de France, iii. Monaxonida (Hadromerina). Arch. Zool. exp. et gén. (8), viii, 1900. . Topsent, E.—Spongiaires des Agores. Résultats des campagnes scientifiques du Prince de Monaco. Monaco, 1904. . TorseNt, H.—Spongiaires provenant des campagnes scientifiques de la Princesse-Alice dans les Mers du Nord. Monaco, 1913. . Wurreieces, '’.—Sponges. Part ii. Scientific Results of the ‘lvawling Expedition of H.M.C.S. “Thetis” off the coast of New South Wales, Mem. Australian Museum, iv, part 10, 1907. 16 Proceedings of the Royal Irish Academy. EXPLANATION OF PLATE L Fic. 1. Mycale (Paresperella) atlantica nu. sp. a, b, styli, x 330; ¢, toxa, x 330; d, ¢, large and small anisochelae x 600; 7, sigmata, x 330. 2. Forcepia fragilis n. sp. a, b, styli, x 140; ¢, tylotum, x 330; d, ¢, two forcipes, x 900 and x 600 ; # isochela arcuata, x 600; g, sigma, x 350. 3. Microciona laevis Bowerbank. a, b, c, ¢, heads of styli of different sizes, x 330; d, small stylus, x 330; £ dermal stylus, x 140; g, head of dermal stylus, x 600 ; A, toxa x 330. Proc. R. I. ACAD., VoL. XXXIV., SEcr. B. PLate I. eZ ab | Etleer E. Barnes STEPHENS.—SPONGES FROM DREDGING EXPEDITIONS. POWAY TE ON THE ATTACHMENT ORGANS OF THE COMMON CORTICOLOUS RAMALINAE. By LILIAN PORTER, M.Sc., From THE HoNAN BronoGcicaL INSTITUTE, UNIVERSITY COLLEGE, CORK, COMMUNICATED BY Sirk BerrRaAM WINDLE, F.RS. PLATES II-IV. Read Junz 11. Published Sepremprr 7, 1917. I.—AIM or RESEARCH. THE aim of the present investigation was, in the first instance, the examination of the so-called “rhizoids” of the corticolous members of the genus Aamalina. It is obvious that these organs are not comparable with the rhizoids of, for example, the Parmeliae, and that the work hitherto published on the subject does not explain all the phenomena connected with the development and spread of these plants. This involved a certain amount of research in the general anatomy of the species under consideration, especially as the investigator is met at the outset by fundamental disagreement or misunderstanding among the leading authorities on the subject. In the course of the work a secondary aim, though one of more practical importance, obtruded itself, namely, an estimation of the amount of damage done by these lichens to their hosts. This amount is generally stated to be negligible, and, in the case of timber trees, probably is so. Evidence accumulated, however, to show that penetration of the host plant by lichen elements is not limited to the dead tissues of the cortex, and demanded more complete examination. II.— History. The earliest detailed work on the anatomy of species of Ramalina is that of D. J. Speerschneider." It explains the characters which distinguish the 'Speerschneider. ‘‘ Mikroskopisch-Anatomische Untersuchung tber Ramalina cali- caris Fr. und deren Varietiiten fraxinea, fastigiata, canaliculata und farinacea” (Bot. Zeitung, 1855, p. 345). R,I.A. PROG,, VOL. XXXIV, SECT. B, [C] 18 Proceedings of the Royal Irish Academy. so-called varieties of R. calicaris, with the exception of R. canaliculata recognized as species, and gives in detail the histological and physical peculiarities of the tissues concerned. Briefly, he recognizes an outer cortex, a gonidial layer and a medulla in all varieties, the relative development of cortex and medulla being in some cases characteristic ; he distinguishes two opposed cortices, an upper and an under; he notes the more or less loose nature of the medullary tissue, and the fact that the gonidial layer is an interrupted one, sometimes penetrating the lichen cortex. In his detailed account of the cortex he asserts that the apical swellings of the longitudinally arranged hyphae are disposed at right angles to the surface, being firmly laid together to form the very smooth surface of the lichen. The cortical hyphae are described as thin tubes with fine lnmen and thick walls, and as being bound together by a cement, the decomposition product of the cell walls; this is stated to be an easily soluble, structureless substance lying between the cells ; no cross walls could be found. In 1860-2 Schwendener’ published his researches on the lichen thallus. He regards the outermost layer, Speerschneider’s “apical swellings,” as the true cortex, strengthened internally by strands of mechanical tissue, which consists of longitudinally arranged hyphae. He recognizes cross walls and regards the “cementing substance” as a part of the thick hyphal wall, supposing that Speerschneider mistook the lumina of the cortical cells for the complete cells and their walls. He distinguishes in all epiphloeodic lichens, whether crustaceous or foliaceous, a basal layer or disc of tissue, free from algal cells, growing within the periderm and serving for attachment and nourishment. This he names the “hypothallus,” remaining in the substratum under the “protothallus.”” In the former term he includes rhizoids of foliose lichens, basal dises of strap lichens and gonidialess hyphal layers of crustaceous lichens. He further states with regard to the mode of attachment, that in Usnea the attachment-point gives rise to rows of hyphae which take root round projections of the bark, thus fastening the lichen to the substratum, and that these attaching hyphae arise from the medullary tissue. In 1894 Crombie, in his monograph, following Nylander,’ distinguishes between species of Ramalina having an amorphous and those having a filamentous cortex, most of the corticolous species coming under the latter heading. He describes the ' Schwendener. ‘‘ Untersuchungen aber den Flechtenthallus ” (Nageli’s Beitrage z. w. Botanik, 2, 4). ? Crombie, J. M. A Monograph of Lichens found in Britain. London, 1894. * Nylander, W. Recognitio monographica Ramalinarum (Bull. Soc, Linn, de Norm. 2 iv, 1868-9). Porter— Organs of the Common Corticolous Ramalinae. 19 medullary layer as arachnoid, becoming sometimes very lax, and the cortical layer as composed either of indistinct cells or of longitudinal conglutinate filaments ; that is, he does not distinguish cortex and mechanical strengthening tissue. Lindau,! 1895, follows Schwendener closely as to anatomy and mode of attachment, but whilst condemning the general use of the term hypothallus, expands the idea. He regards it as a structure of secondary origin, the hyphae of which are able to penetrate between the periderm cells, forcing apart the cell layers, but unable to bore through the walls. The occurrence of definite cavities and cell remnants connected with the basal dise is attributed to the chemical action of atmospheric agents, and he suggests that the hyphae are able to assimilate the decomposition products of the cellulose, which are thus rendered available as nourishment for the whole organism. He figures and explains the basal disc in Lvernia prunastri as clearly cut off from the thallus itself by a narrow zone of brown-coloured hyphae, which is to be regarded as the outermost layer of the lichen thallus. This hyphal cylinder is open below, and the medulla of the lichen is directly connected with the hyphae of the basal disc. On this appearance he bases his view that the dise is secondary in origin, and supports it by the fact that broken-off, isolated parts of the periderm, which are surrounded by hyphal masses, move gradually outwards. He shows, too, in Hvernia that sucker-like outgrowths arise. A branch of the thallus may attach itself to the’ bark of the .twig on which the plant grows, or to that of a neighbouring one. If the connexion with the mother plant is accidentally torn, an independent individual arises, the origin of which is not obvious later. He refers elsewhere to the thickly-packed, alga- containing lobes of the cortex occurring in connexion with the thallus itself, but dismisses them as non-typical in structure, mere occasional forms of attachment, having no further meaning. Reinke’ includes in his account of the comparative morphology of lichens a review of the Ramalinae. He refers to Schwendener’s work on the anatomy of this genus, but disregards the mechanical tissue, indicating merely the presence of a thickly-woven cortex, a loose medulla, and a gonidial layer between them. Darbishire,s in his account of the Ramalinae of Schleswig- 1 Lindau, G. ‘“ Lichenologische Untersuchungen, I.,. Uber Wachstum und Anheft- ungsweise der Rindenflechten.”’ Dresden, 1895. 2 J. Reinke. ‘‘Abhandlungen iiber Flechten” IV (Pringsheim’s Jahrbiicher f. w. Botanik 28, 1895). 30. V. Darbishire. ‘‘ Kiner Abhandlung tiber die Naturgeschichte der Einheimischen Flechten,” with R. v. Fischer Benzon’s Die Flechten Schleswig-Holstein. Kiel, 1901. [C2] 20 Proceedings of the Royal Irish Academy. Holstein, distinguishes merely cortex, gonidial layer, and medullary tissue ; he describes and figures the ruptures of the cortex of R. fraxinea, where the medullary tissue comes to the surface, and names them “Atemporen,” i.e. breathing pores; he further regards the attaching organs as the portion first formed. In 1901, also, Bitter,' in his account of the variability of some foliose lichens, describes &. obtusata Arn. as having two opposed cortices, scattered groups of gonidia, and a loose mesh of medullary tissue. He observes the effect of light on the cortices as producing gaps in the cortical tissue on the under surface, but rarely on the upper. He cites similar occurrences in Z&. dilacerata Hoffm., and in &. minuscula, a rare British species; he refers to gaps in 2. fraxinea as piercing the thallus. He further points out in reference to Lindau’s account of the sucker-like reproduction of Lvernia prunastri, that in the allied but prostrate species 2. furfuracea, the method is common, Brandt,’ in 1906, investigated the anatomy of many species of Ramalina, He discusses previous work, dismissing that of Speerschneider as primitive in method, and regarding all failures to distinguish between cortex and mechanical tissue as regrettable oversights. He explains that his material is dried and treated with potash to dissolve out the usneic acid between the hyphae, neutralized with acetic acid, and finally stained with alcoholic solution of iodine. By these means he finds pseudoparenchymatic tissue in the mechanical strands, consisting of more or less isolated hyphae enclosed in dark walls—an appearance which will be discussed later. He also regards the tissue of the basal dise as secondary in origin, but does not give details of its structure nor of his reasons for the view. His account of the varying proportions of cortical and other tissues present in the different species, in the light of Bitter’s work on the effect of external influences, may possibly be unnecessarily detailed, the differences being physiological rather than specific. Zahlbriickner* also follows Schwendener in distinguishing an inner cortex of mechanical tissue in which the hyphae run more or less longitudinally, and a true cortex in which they are more or less at right angles to the longitudinal axis; he regards the inner cortex as frequently cut into supporting “pillars” by out- growths of the medullary tissue towards the surface. Fiinfstiick,‘ in his general account of lichens, refers to the thickly-woven ! Pringsheim, 36. 1901. 2 Brandt, Th. Beitrage zur anatomische Kenntnis der Flechtengattung Ramalina (Hedwigia, 45. 1906). ’ Zahlbriickner. Engler and Prantl’s Natiirlichen Pflanzenfamilien—Flechten Teil I, p- 220. ‘ Fiinfstiick. Engler and Prantl’s Natirlichen Pilanzenfamilien—Flechten Teil I, p- 20. Porrer— Organs of the Common Corticolous Ramalinae. 21 strands of the Ramalinae as possessing only a slight capacity for water con- duction. He further states the view generally held that rhizoids of corticolous lichens never reach the living tissues. II].—MarertaL and Meruops USeEp. The species of Ramalina used in this investigation were the commoner corticolous ones—2. calicaris, fraxinea, farinacea, fastigiata, (populina), and pollinaria, i.e. the closely allied species of the section Huramalinae Stizbg. and the sub-section Compressiusculae Wainio, characterized by the flattening or bilaterality of the thallus segments. All the above species agree in being more or less longitudinally striato-nerved, and possessing a filamentous cortical layer (Crombie); the medulla gives no reaction on treatment with potash. LR. calicaris and fraxinea do not give rise to soredia, and the latter is distinguished by its broader, flatter thallus, largely rugose receptacle of the apothecia, and its curved spores. Rh. fastigiata is distinguished, especially from young states of R. fraxinea, by its more caespitose habit and its numerous apothecia, which are peltatosessile rather than distinctly stalked; its spores are curved. R. farinacea is typically more or less covered with whitish, round or oval soredia, especially towards the tips of the laciniae, the bases being frequently smooth; its apothecia are rare. R. pollinaria differs from £. farinacea in that the laciniae of the latter are stiff; those of the former tend to be flaccid and inflated. Its soredia are white, farinose, and scattered; its apothecia are rare. The morphology of the above species was studied in the fresh condition, abnormal and sub-specific forms being discarded. Comparisons were made with the collections in the Herbaria of the University of Manchester and of University College, Cork, by kind permission of the curators, Miss G. Wigglesworth and Miss B. E, Duke, and at Bangor with the collection of Mr. Griffith,’ to whom my thanks are due. The anatomical investigations were carried out as far as possible on the fresh material, comparisons being made with the dried material, preserved material and specimens treated with various reagents. Throughout the observations recorded potash was used with caution, as its clarifying properties are neutralized to a great extent by its action on the walls of the hyphae, and especially by the swelling it causes in the walls of the periderm, 1 Griffith, J. Flora of Anglesey and Carnarvonshire, Bangor. 22 Proceedings of the Royal Irish Academy. When reagents were used with hand sections the method of Lindau was adopted, zc. treatment with alcohol to drive out air, transference to chloral hydrate, and finally, after preliminary examination, treatment with chlor- zinc-iod or iodine solution; this method results in a limited and uniform swelling of the section. For microtome sections the thallus, with a squared block of the stem or twig to which it was attached, was fixed in alcohol, and then transferred slowly from pure alcohol to pure xylol (25 per cent. aleohol, 50 per cent., 100 per cent.; 5 per cent. xylol, 10, 25, 50, 70, 100 per cent.). Penetration of the tissues by paraffin wax dissolved in xylol is a slow process which cannot safely be hastened in any way, since the tissue of the periderm tends to burst, making complete sections impossible, and creating artificial lacunae. For differentiating stains chlor-zinc-iodine was used; it gave better effects when used after potash. Fuchsin gave good results when varying thicknesses of cork cell walls were to be examined, or to show up the path of a hypha between the walls. In cases where hyphae apparently passed through a cork cell wall the illusion is best dispelled by use of Hoffmann’s Blue. TV.—DESCuKIPTION. 1. General anatomy of the thallus, Fig. 1 represents a longitudinal section of the thallus of 2. farinacea, illustrating the closely packed longitudinal hyphae of the general cortical tissue (c); they turn outwards to the surface at their apices, but hardly at right angles to the axis; the general effect may, however, be roughly stated as “more or less at right angles.” Further, the assemblage of these apices is, in the specimen depicted, scarcely sufficiently developed in comparison with the rest of the cortex to justify its being regarded as a tissue sui generis. In specimens obtained from more exposed localities a broader expanse of apices does undoubtedly occur, but is correlated with a greater width of longi- tudinally arranged tissue; a similar correlation is observable in the lighted and unlighted surfaces of the thallus of &. frazinea. This contrast between the cortices of the two surfaces is not so marked in thalliof 2. farinacea, since in that species there is greater tendency to cylindrical development of the laciniae, In the figure we have an example of the emergence of the gonidial (g) and medullary (m) tissues at the surface, but whether the phenomenon is con- nected with the development of soredial structure or an “atempor” is irrelevant to the present investigation. Portrr—Organs of the Common Corticolous Ramalinae. 23 The grouping of the apices of the hyphae into lobes or bundles should be noted, as it is apparently continued for a short distance into the general cortex. Fig. 2 is a cross section taken near the base of a bushy specimen of R. furinacea just above the surface of the periderm of Oak. The white patches (ec) are pillars and convolutions of cortical tissue, i.e. longitudinal hyphae. The dark internal portions consist of loosely-woven medullary tissue, edged here and there with a few gonidia. Externally and between the lobes are masses and layers of varying character, in places distinctly soredial or even gonidial, but as a rule indefinable. ‘The figure resembles Brandt’s figures of R. Curnowii.! Whether each hyphal mass is continuous with one lacinia would be hard to decide, but they are frequently separate and con- tinuous for some distance into the periderm. Fig. 3 represents a portion of the transverse section of the thallus of Rh. farinacea, showing the more or less horizontal direction of the apices (@) ; the cut ends of the vertical hyphae, interspersed with other hyphae running in various directions ; and, internally, the gonidial and medullary tissues. 2. Attachment of the lichen, and anatomy of the hypothallus. Tn fig. 4 we have the longitudinal section of R. farinacea on Hawthorn. This is very similar to the figures given by Lindau representing the attach- ment of strap and fruticulose lichens to their hosts. We have here the apparently sudden end of the portion of the thallus above the periderm (p), whereas the hypothallus or basal dise consists of hyphae or hyphal masses scattered among the periderm layers. Here, too, may be seen the wedge-like, disintegrating action of these hyphae on the cork cells. Figs. 5, 6, 7, and 8 represent the paths of hyphae between the cells and across the gaps in the periderm tissue. 5, 6, and 7 are portions of R. furi- nacea on Hawthorn, 8 is from a transverse section of &. ecalicaris on Sycamore, showing the spread of hyphae outwards from the basal disc. In all these sections 4-8 the attaching hyphae appear inconsiderable in bulk and power compared with the thallus to be supported, and with the size of the gaps in the periderm. In some cases the latter discrepancy is due to the growth of the periderm after penetration by the hyphae. But in all these sections the two points of interest are the bridging of gaps by single hyphae and the close adherence of the hyphae elsewhere to the walls of the periderm cells, such cells frequently presenting a crushed or collapsed appearance. ' Brandt, loc. cit., p. 152, and Plate VI, figs. 1 and 2, 24 Proceedings of the Royal Irish Academy. In the next series of figures we have the explanation of the apparent discrepancy between the size of the basal disc and that of the thallus which depends on it for resistance to wind action. In fig. 9, a longitudinal section of #. calicaris on Sycamore, we have a well-developed specimen of many years’ growth. It has one definite main “hapteron” (hk), consisting of longitudinally arranged, intertwined hyphae continuous with the similar cortical hyphae (c) rather than with the loose tissue of the medulla (m) which here ends above the level of the basal disc. From the main hapteron arise lateral branches of hyphal tissue penetrating the periderm in all directions. In sections of older thalli, fig. 10, the medullary tissue extends a hoc distance below the surface of the periderm, and is seen to merge gradually into the more compact cortical tissue. Possibly its appearance below the surface level of the periderm is due to the growth of the latter round the base of the thallus proper. In fig. 11 of 2. farinacea on Oak we have a branched system of haptera with islands of periderm tissue and cell-remnants scattered among the hyphae. The layers of cork cells are obviously crushed and distorted. The lichen thallus above the periderm gradually merges into a layer of algal cells, fungal hyphae, bark debris, &c., which spreads for some distance over the surface of the stem. In fig. 12 this layer (/) is especially well developed. Figs. 12 and 13 are two of a series of transverse sections taken between a young specimen of R. calicaris and its parent plant, *3 cm. distant on a twig of Prunus spinosa. It will be seen in fig. 12 that though a well-formed hapteron occurs in the outer layers of the periderm (p), the gaps containing smaller hyphal masses are most frequent towards the bast /+) and cambium (k), even affecting the wood (w). In fig. 14 we have an example of the spread of a young Ramalina in the spine of a Hawthorn. Considering the hard, smooth surface of such a spine, it is clearly probable that the younger plants arose from hyphal masses within the periderm, rather than from sporal or soredial developments on the surface. 3. Action of the lichens on the host plant. In fig. 15, representing a young #2. frazinea on Alder, the centre of the twig is completely decomposed, and the outer part below the lichen shows signs of decay. In 16 and 17 we have sections of the edge of the hollow nearest the lichen, showing the state of the affected tissue and the extension of branched and septate hyphae into the open space. It was not found possible to trace any Portrr-— Organs of the Common Corticolous Ramalinae. 25 hypha or mass of hyphae directly from the lichen to the centre, and the possibility remains that the hyphae figured belong to some species of parasitic fungus ; but, on the other hand, the section represents a state of affairs too frequent to be the result of the chance co-operation of two sets of mycelia. Fig. 18 represents hypertrophy of the periderm and erosion of the wood. Here the host is Alder, and the lichen probably £&. pollinaria, as well-developed plants of this species were frequent on this and neighbouring twigs. Not only does the typical periderm tissue develop as a cushion round the base of the lichen, but, besides the hyphal masses penetrating the external tissues, we have tongues of tissue consisting of polygonal, pitted cells (1), apparently eating their way into the otherwise normal wood (7), as in fig, 19. Fie. 20 gives in detail the character of these ingrowths. V.— Discussion OF RESULTS. With regard to the structure of the cortex of these lichens, the figures obtained hardly seem to justify the distinction between cortex and mechanical tissue. For convenience the terms inner and outer cortex might be retained, with the reservation that the one is a mere continuation, with more or less change of direction, of the other. The swollen appearance of the apices described by Speerschneider is probably due to his use of caustic potash, since other less violent clearing agents, such as chloral hydrate, do not produce this effect. The pseudoparenchyma described by Brandt has been alluded to, and may certainly be more or less satisfactorily obtained by his method. If, however, fig. 3 be examined closely, it will be noted that the cortex consists of hyphae cut in all directions, the majority transverse. On treatment with potash these last appear as isolated cell sections, whilst the interwoven hyphae form more or less complete enclosures round them. On addition of acetic acid this appearance becomes still more sharply defined, and on staining with alcoholic iodine the hyphae become coloured so that the contents of the cell sections show up clearly, and the interwoven hyphae apparently form cell-walls around them. In the intraperidermal extensions of the cortical cells it is almost im- possible to induce this parenchymatic appearance, and this may be attributed to the fact that here the hyphae are more rarely interwoven, the main direction being longitudinal. With regard to the action of potash as a clearing agent, it should be remembered that this alkali is a solvent of usneic acid, which is apparently the most common of lichen acids. In the case of R. thrausta it occurs as minute crystals in the cortex, giving to that tissue its yellowish-green colour, R.LA. PROC., VOL. XXXIV, SECT. B. [D] 26 Proceedings of the Royal Irish Academy. though only present to the extent of -3 per cent. as shown by Zopf.! In the ease of R. farinacea it is present to the extent of 25 per cent.,in R. pollinaria ‘16 per cent.; but undoubtedly of the three species, 2. farinacea is normally the lightest in colour, apart from the whitish, farimaceous appearance given by its soredia. Further, if the acid be present in crystalline form it can hardly appear as a structureless cement between the hyphae. Its presence is commonly regarded as substantiated if the addition of caustic potash gives a yellow colouration, but this generalization should be regarded with suspicion, as a similar colour is given with potash by ramalic acid in KR. pollinaria, by cetrarie acid in species of Cetravia and Cladonia, and probably by others. Again, it should be noted that these acids are obtained and investigated after extraction from dried and crumbled lichens by means of hot ether (Zopf)! or milk of lime (Stenhouse, Hesse),* and cannot therefore be regarded with any certainty as occurring naturally in the forms described, except in the few rare cases exemplified by 2. thrausta. The excreted substances, acting as cements, are therefore probably of the nature of lichenin, the gum-like substance occurring in Cetraria islandica, the Iceland Moss; they become mucilaginous and semi-transparent in water, as noted by Speerschneider, but are hydrolyzed readily with acids, and in some cases with alkalis, forming in some cases sugars, e.g. glucose from the extract from Evernia prunastri, mannite from Xanthoria parietina. This gummy character would account for the adhesion of the hyphae to the cell-walls of a gap in periderm tissue, enabling them to absorb the decomposition products of the cellulose, according to Lindau’s theory (Untersuchungen, I, p. 57). In this case the basal dise must be regarded as an organ of nourishment as well as of attachment, an important conclusion, involving the partial independence of the fungal hyphae of their algal companions and em- phasizing the symbiotic, rather than the parasitic, nature of the combina- tion (Finfstiick, p. 15). With regard to the attachment organs themselves, three points distinguish those of the Ramalinae investigated from those of Usnea and Hvernia as described by Schwendener and Lindau. In the first place there is no limiting zone, the tissues of the thallus above and below the surface of the periderm being continuous. In the second, this tissue is cortical, consisting of thickly-woven strands which only become loose masses at their apices ' Zopf, W. Zur Kenntnis der Flechtenstoffe vii. Annalen der Chemie, 313, 227, and 300. * Hesse, O. Uber einige Flechtenstoffe. Annalen der Chemie, 284, Porrer Organs of the Common Corticolous Ramalinae. 27 or periphery. In no case has the medullary tissue been traced into the peridermal gaps. Finally, in the Ramalinae there is a definite “hapteral system ”—the term is used tentatively—consisting of one or more main haptera piercing the periderm, with lateral branches in all directions. The whole system resembles the “cortex roots” and suckers of Visewm album, the Mistletoe, but has not their parasitic function to such a marked degree. The periderm tissue bordering the main and lateral haptera is crushed into confused, dark-coloured layers, and fragments of cells lie scattered among the hyphae. The lateral haptera taper to a point which is lost among the periderm cells, apparently dwindling into small bundles of hyphae which are capable of changing their direction of growth, and of branching and increasing in diameter. They exert a wedge-like action on the cork layers, as described by Lindau. : The question therefore arises as to whether the bushy masses in which the Ramalinae often occur are the result of direct development of spores or of soredia, or of “runners” from a parent plant; and further, in the last case, whether the effective runners are superficial or intraperidermal. In this connexion it is interesting to note the view of Bornet,' that the algae of hypophloeodic lichens are already present when the fungus begins to immigrate, and the opposed view of Frank,’ that from a germinating spore a protothallus grows into the periderm and only later is colonized by immigrating algae. The sucker-like outgrowths of Hvernia prunastri as described by Lindau have been mentioned above. This method of reproduc- tion is undoubtedly very common with the Ramalinac, but probably not so much so as that of the stolon-like forms. Assuming that the plant reaches its host from the outside, the bundles of hyphae grow out and penetrate between the periderm cells. The best points of entrance are where a crack in the bark leads to a weak place in the tissue, or a lenticel. Here, too, is a suflicient supply of water or moisture for the growth of the hyphae; consequently the hyphal bundle increases in length and diameter, and crushes aside the dead cells of the bark. The surface of these hyphal bundles being somewhat of the same character as that of the lichen cortex, the lateral haptera are formed by the intrusion of one or more hyphae between the horizontal layers of periderm, the exact 1 Bornet, E. Recherches sur les gonidies des Lichens. Ann. des Se. Nat., 5 ser. Bot. 1873, xvii, p. 45. 2 Frank, A.B. Uber die biologischen Verhiiltnisse des Thallus einige Krustenflechten. Cohn’s Beitr. 1877, ii, p. 123. [D 2] 28 Proceedings of the lioyal Irish Academy. stratification of which facilitates the splitting process. This power of the hyphal tissue has been demonstrated by Lindau in his account of the develop- ment of suckers. “The events which lead to the attachment of the thallus are as follows :—Out of the superficial layer spring hyphae in thick bundles of parallel threads which reach to the surface of the substratum, and here by any crack leading into the loose tissue of the upper periderm layers act exactly as the hyphae of the basal disc... . .” Such a system, especially when composed of numerous haptera, is capable of keeping pace with the development of the bark, and the thallus need not be torn off by wind action when the outer periderm is quite dead. Probably therefore the theories of Bornet’ and Frank* are not mutually exclusive, Bornet’s theory being supported by the reproduction of Ramalinae by the sucker and stolon methods. Protococeus cells are very frequent on the surfaces affected, and in two examples, one of &. Marinacea and one of NXanthoria parietina, were found in the periderm some distance below the surface, apparently unconnected with any hyphae or hyphal mass. No clear case has yet, however, arisen to prove conclusively that a young plant is directly or solely connected with the parent plant by means of an intraperi dermal hyphal strand. In most cases the evidence consists of the existence of such strands running more or less horizontally along the twig, a state of affairs comparable with the “Flechtenmycel” of Zukal, which he defines (ex Finfstiick, p. 8) as “ hyphal complex usually arising from an old lichen thallus, which often permeates the substratum for a foot’s length, and gives rise at single points to new thallus formations, eg. in Peltidea venosa, Solorina saccata, Diploschistes scruposus, Xanthoria parietina, Cladoma macilenta, &c.” The surface of a twig is frequently covered by a more or less continuous gonidial layer which is too uniform to be considered as soredial in character, though possibly fragments rounded off in times of drought may act as soredia, or under other adverse circumstances may produce the familiar Lepraria forms. There is, as yet, no evidence, in the case of Ramalina, of direct vertical connection between this superficial layer and the intraperidermal hyphal strands, except where a plant-outgrowth occurs. The conjunction of these strands, at a favourable crack or lenticel, may be the determining condition for the development of a new plant. Again, Xanthoria supplies suggestive evidence. A specimen growing on a twig of Hawthorn with ' Bornet, E. Recherches sur les gonidies des Lichens. Ann. des Sc. Nat., 5 sér. Bot. 1873, xvii, p. 45. ? Frank, A. B. ‘‘ Uber die biologischen Verhdltnisse des Thallus eniger Krusten- fletchten Cohn’s Beiter,” 1877, ii, p. 123. Porrer—Organs of the Common Corticolous Ramalinae. 29 four annual rings had for length of lobe from centre *25 inches, and at a distance of ‘37 inches a small specimen of the same species occurred on a branch showing three annual rings, the intervening surface being covered by a thin layer of gonidia containing hyphal tissue and the periderm being traversed horizontally by hyphae. On the other hand, this superficial layer has been seen in connexion with Xanthoria, covering a lenticel without sending downward growths into it. Injurious Effect of Lichen Growth on Trees. Though it is usually asserted that lichen growth does not to any great extent injure trees from the point of view of the forester, there is no doubt that it is regarded unfavourably by the arboriculturist, and in particular, the fruit grower. Even the forester may admit that boughs broken off in a storm are often covered by a luxuriant growth of lichens, but there is no direct evidence that such growth renders boughs more liable to damage by wind. The indirect evidence on which Lindau bases his assertion that the effect of the hyphae is directly injuzious may be summarized as follows :— 1. They block the lenticels. . Packing tissue of lenticels disappears. . Cork layers are torn apart horizontally. Apart from lenticels, vertical cracks are forced open. oP oo Lo Cell walls are frequently torn across, especially in Oaks. 6. Atmospheric decomposition of the cell walls causes decrease in thick- ness of cork cell walls from within outwards, thus rendering the cork less impermeable. In fig. 12 we have evidence of more detinitely harmful action, as here the hyphae oceur in periderm, bast, cambium, and even the youngest cells of the wood. That, in this case, the hyphae are not those of a parasitic fungus is proved by the fact that they did not occur in these positions except between the young plant and its supposed parent. A parasite could hardly be limited to this short distance. This is, as yet, the most striking case observed in which the hyphae were found inside the periderm, and occurs in Prunus spinosa, the Sloe, a shrub of very leisurely growth. Hartig' remarks that “the slower the growth of a 'Hartig, R. Diseases of Trees: Trans. Somerville and Marshall Ward, London, 1894, p- 5d. 30 Proceedings of the Royal Irish Academy. tree the more slowly do the outer cortical layers die, and so much the more suitable are the conditions for the growth of lichens.” He compares the absence of lichens on Beeches grown under favourable conditions, i.e., on calcareous soil, with their abundance on the slow-growing Beeches of sandy soils. Lindau in his section on lichens, in his edition of Sorauer’s Plant Diseases,’ states that it is impossible for the hyphae to penetrate within the periderm tissues except possibly by means of cracks or fissures already present. He had not observed any cases of the presence of hyphae in living tissues in his researches on a list of species which includes those of Ramalina. He appears, however, to have used as material chiefly the larger species of tree for his researches on fruticulose lichens. Further evidence must, however, be sought before his conclusion can be disproved, founded as it is on many investigations and much knowledge of lichen-gitowth. There is, however, this further consideration. He differentiates between the harmful actions of lichens on fruit and on forest trees. In the former case he points out that the lichen holds water, and so keeps the bark moist, hastening its decomposition, and at the same time offering itself as a shelter to noxious insects. In the latter he considers that wealth of lichen-growth indicates unfavourable conditions, dampness of situation, exclusion of light, poor soil ; and that these are the causes of the death of such trees rather than the lichens themselves. On the other hand, if conditions are so bad that the rate of annual apical growth is reduced to a minimum, the lichen may cover the tips of the branches and cause suffocation. The remedy is here obviously to allow the entrance of light and wind. In the case of forests on a large scale these statements hold, but for the small scattered woodlands of the British Isles it is more frequently found that the maximum lichen-growth is on the side of the wood, and of the trees in the wood, which is exposed to the prevailing wind, ie., the moisture-laden south-west. The light factor is here not of such great importance as it would be in a wood of great extent- Judging, indeed, from the luxuriant growth of lichens on the exposed sides of stunted Oak, Birch, Hawthorn, and Sloe around our coasts and on the highest wooded portions of our mountains, it would seem that moisture and slow growth of the host were the important factors. The presence of salt in the moisture is apparently not harmful to the lichens, though wind-blown sand is undoubtedly so, the action probably being a purely physical one. 'Sorauer, P. Handbuch der Pflanzenkrankheiten ii, 1908, p. 484. Porter—Organs of the Common Corticolous Ramalinae. 31 The remedies for lichen-growth on fruit trees may be summarized as follows :— 1. Scraping off the outer bark scales and painting with lime-wash. 2, Spraying with Bordeaux mixture. 3, Spraying with 5 per cent. soda solution to which 3 to 4 per cent. carbolic acid has been added. If the percentage of acid is increased, the spraying must be done in winter or the buds will be injured, though the increase has the merit of removing scale and other insects. VI.—SUMMARY OF CONCLUSIONS. 1. The thallus of the Ramalinae consists of cortex, gonidial layer, and medulla. 2. The cortex consists of longitudinally arranged hyphae whose apices turn outwards to the surface. 3. Differentiation of the inner and outer cortices, and pseudoparenchy- matic appearance of the former may be produced by suitable reagents. 4. The attachment organs are strands of closely woven hyphae, longitudinally arranged, and continuous with the eortical tissue. 5. They penetrate the periderm by cracks or lenticels, and by wedge-action cause extensive splitting. 6. They give rise to branches in all directions. 7. From these horizontal branches, or from the superficial soredial layer, or from both acting in conjunction, arise new plants. 8. These organs may penetrate the living tissues as far as the new wood, causing direct injury to the tree. 9. They also cause hypertrophy of the peridermal tissue, and erosion of the wood by ingrowths of hypertrophied tissue. In conclusion, the author’s thanks are due to Dr. O. V. Darbishire, of Bristol University, for proposing the subject under consideration, and for the loan of otherwise inaccessible literature; to Professor I. W. Phillips, of the University College of North Wales, Bangor, for advice and encourage- ment in the practical investigations; and to Professor M. Hartog, of University College, Cork, for criticisms and suggestions as to the arrangement of the work. 32 bo 9. 10. Proceedings of the Royal Irish Academy. BIBLIOGRAPHY. . SPEERSCHNEIDER, D, J.—Mikroskopisch-Anatomische Untersuchung iiber Ramalina calicaris Fr. und derem Varietiten /fraxinea, Sastigiata, canaliculata und farinecea. Bot. Zeitung, 1855, p. 345. . SCHWENDENER, S.—Untersuchungen tiber den Flechtenthallus. Naegeli’s Beitrage z.w. Botanik, 2, 4. . CromBiz, J. M.—A Monograph of Lichens found in Britain. London, 1894. . NYLANDER, W.—Recognitio monographica Ramalinarum. Bull. Soe. Linn. de Norm. 2, iv, 1868-9. . Linpau, G.—Lichenologische Untersuchungen, i, Uber Wachstum und Anheftungsweise der Rindenflechten. Dresden, 1895. . RemxKe, J.—Abhandlungen iiber Flechten, iv. Pringsheim’s Jahrbicher f. w. Botanik, 28, 1895. . Darsisuire, O, V.—Einer Abhandlung tiber die Naturgeschichte der Einheimischen Flechten, with KR. v. Fischer-Benzon’s Die Flechten Schleswig-Holsteins. Kiel, 1901. . Brrrer, G.—Uber die Variabilitat einiger Laubflechten and tiber den Einfluss ausserer Bedingungen auf ihr Wachstum. Pringsheim’s Jahrbiicher f. w. Bot. 36, 1901. Branpt, Fu.—Beitrage zur anatomische Kenntnis der Flechtengattung Ramalina. Hedwigia 45, 1906. ZAHLBRUCKNER, A.—‘“ Flechten” in Engler und Prantl’s Natiirlichen Pflanzenfamilien, Teil i.* B. Specieller Teil, 1907. . Finrstick, M.—Jvid. A. Allgemeiner Teil. 2. Grirriru, J. E—Flora of Anglesey and Carnarvonshire, Bangor. 3. Zopr, W.—Zur Kenntnis der Flechtenstoffe, vii. Annalen der Cheiie, 313, 297, and 300. . Hesse, O.— Ueber einige Flechtenstoffe. Annalen, 284. 5. Bornet, E.—Recherches sur les gonidies des Lichens. Annales des Se. Nat., 5 sér. Bot. 1873. . Frank, A. B.—Ueber die biologischen Verhaltnisse des Thallus einige Krustenflechten. Cohn’s Beitrage zur Biol., 1877, ii, . Hartic, R.—Diseases of Trees. Trans. Somerville and Marshall Ward. London, 1894. . Soraver, P.—Handbuch der Pflanzenkrankheiten, ii. 1908. Proc. R. I. AcAp., VoL. XXXIV, Secr. B. PLATE II. PORTER.—ATTACHMENT ORGANS OF RAMALINAE. i” a ; A io I 6. 7 Seen | Oe ” os oS 7 ¥ 7 ad > we a 9) PLATE III. Proc. R. I. ACAD., VoL. XXXIV, Sect. B. ATTACHMENT ORGANS OF RAMALINAE, PORTER. PLATE IV. Proc. R. I. AcAD., VoL. XXXIV, Sect. B. {) ed 044 ita) a mJ PORTER.—ATTACHMENT ORGANS OF RAMALINAE. [ 83 ] III. THE EXPLORATION OF CASTLEPOOK CAVE, COUNTY CORK, Being the Third Report from the Committee appointed to Explore Trish Caves. sy R. F. SCHARFF, B.Sc., Po.D.; Proressor H. J. SEYMOUR, B.A., B.Sc, F.G.S.; anp E, T. NEWTON, F.R.S. Puates V.-VII. [Read Junr 25, 1917. Published January 9, 1918.] CONTENTS. Secr. PAGE 1. Iyrropucrion. By R. F. Scuarrr, < : : © GB) 2. Grotocy. By H. J. Seymour, : : c . 46 3. Anima Remains (except Birds), By R. F.Scuarrr, . 49 4. Brrp Remains. By E. T. Newton, : . 5 a GY 5. Summary or Resunts. By R. F. Scuarrr, . : - 68 1.—INTRODUCTION. IN our second report? some of the caves of Co. Clare, and their contents, were described. The present report deals with a cave considerably further south, namely, in Co. Cork. Situated about two and a-half miles north of Doneraile, in the townland of Castlepook South, it is easily accessible from an old quarry by a low tunnel. Mr. Ussher’s attention was first drawn to it by Lord Castletown, on whose estate the cave is situated, and who afforded Mr. Ussher every facility for its exploration. Myr. Ussher at once recognized that this investigation was likely to yield valuable results; and having had full charge of similar operations in previous years, he commenced the excavation of Castlepook cave on the 1st September, 1904. Unfortunately, 1“The Exploration of the Caves of County Clare.” Trans. R. I. Acad , vol. xxxiii, Sect. B, pt. 1. 1906. 2 A short description was read before the British Association meeting in Dublin, and published in the Report of the Association of 1908. R.J.A. PROC., VOL. XXXIV, SECT. B, (4) 34 Proceedings of the Royal Irish Academy. he is no longer with us, and it issad to think that this report has to be issued without bearing the name of the one who bestowed such painstaking and valuable work on the undertaking. It is now four years since we have had to mourn Mr. Ussher’s death, and I may safely state that it will be impossible to find anyone to devote more care and industry to work of that nature. Cave exploration was with him a labour of love; and, in spite of his advancing years, he willingly exposed himself to the greatest hardships, and even dangers, in order to gain the knowledge which was procured by his unremitting activity. The Committee are much indebted to Professor H. J. Seymour, of University College, Dublin, for his valuable services in connexion with this exploration. Mr. E. T. Newton kindly undertook to name the bones of birds obtained in the cave, while Mr. Hinton placed his special knowledge of the rodents at my disposal, and was most helpful in determining the lemming remains. ‘To all these gentlemen the Committee wish to express their thanks. No less than 343 parcels of bones were sent to Dublin from the cave. As each of these contained on an average about 100 specimens, which had all to be sorted, cleaned, and catalogued, this alone entailed much labour. Of these 34,000 bones, only the small fraction of five or six thousand could be accurately named. By far the greater number were too fragmentary for determination. The entire collection of animal remains and rocks procured during the cave exploration has been deposited in the National Museum of Ireland. I have to acknowledge the valuable help given by Dr. Smith Woodward and Dr. Andrews during my examination of some of the cave bones at the British Museum. The present entrance to Castlepook cave' is situated about 300 feet above sea level (Plate VI). It is close to the road in a disused quarry, and not far from a stream that falls into the Bregoge. The stream flows southward from the mountains, and carries with it the sandstone detritus, filling all the river valleys of the district. To the north and west of the cave is a marshy flat, probably the bed of a former lake, and this is flooded at times by the Castlepook stream. A series of low crags here extends in east-westward direction, and in partially quarrying the stone away for building purposes one of the former entrances to the cave was thus laid bare. Mr. Ussher refers to others, but these are blocked by detritus near their orifices. Some of these may at times have been used as stables forcattle. Before Mr. Ussher explored the cave it had apparently been entered within recent years, and 1 | retain the original name as given above rather than that of ‘‘ Mammoth Cave,” a term more recently applied to this cave by Mr. Ussher. ScCHARFF, StyMour, AND Newron—Castlepook Cave. Si) ) ) ] most of the halls and galleries could have been examined by anyone possessing the requisite courage and tenacity. Some of them could not have been visited by man before Mr. Ussher made them accessible by excavation. He penetrated into the cave to the distance from the entrance of about 400 feet, but the portion he explored may only be a small fraction of the whole. By removing the earthfalls and rocks now obstructing many of the passages new galleries may be discovered. It is doubtful whether the cave ever was suitable for human occupation. The early inhabitants of Ireland frequently lived in caves, but they selected such that were dry enough in winter; and they neyer penetrated very deeply into the recesses of a cave, but lived in close proximity to the entrance, where they still had the advantage of a certain amount of daylight. : Mr. Ussher noticed that sand was present in all the halls and galleries (except where it has been washed away), and that the great harvest of animal remains occurred whenever it was found to be of a deep red or yellowish- brown colour. Bones were frequently met with on or near the surface of a bed of sand. Where a stalagmite floor had been formed on this bed of sand, and the sand beneath it had been washed away, bones were sometimes seen adhering to the bottom of the stalagmite. In other places the upper sand was rich in bones, the lower barren. The following is a short description of the cave, written by Mr. Ussher and sent to me a couple of years before his death :— “The ENTRANCE GALLERY (Plate VII, 1 to 2) is one of a series of orifices in the base of the crags. Four of them are open caves, which run into the rock from north toe south, and each of them exhibits the features of tunnel and shelf. We excavated one of these, the Goat-house, and found a stalagmite floor which had been undermined by the washing away of the bed on which it had stood; and a loose, pale, barren sand, with cobbles of sandstone in it, had been intruded into the hollow thus formed. In the Entrance Gallery we found sandstone cobbles in limestone rubble on the surface, and beneath them a thick floor of granular stalagmite about three feet below the shelf, with a Bear’s humerus embedded in it beneath, and under this a washed-out hollow into which pale, barren sand had entered. In sinking a hole for the door frame, near 2, some bones of Bear and Reindeer turned up; but in the Entrance Hall (2) and Cloak-room leading east from it was deep, barren sand, only yielding fox-earth bones. This Cloak- room ends in an earthfall probably covering another entrance. We follow a diagonal passage to 3; all. these have traces of the shelf, but no stalagmite un situ. E 2 36 Proceedings of the Royal Irish Academy. 2 Before coming to 3 we cross the line of the Little Narrow Gallery, which affords a beautiful example of the swelling tunnel above the deep . trench, and of the undermined walls of the latter hollowed away beneath. ‘The line of this gallery crosses the Hyzna Hall, and one side of its tunnel and shelf may be seen in the south part of the latter. It is then interrupted by a break-up of the roof, but is continued in the deep gallery called the Abyss. The Lyana HA (3a to 4 and 5), so called because the jaws of the Hyena were found here first, but not to be confounded with Hyena Land. This hall is over 40 feet x 30 feet, and is formed of seven parallel galleries, the north ends of which may be seen in their vaultings ; but their dividing partitions have collapsed, and these, together with fallen roof-blocks, formed a pile filling the body of the hall in 1904, when we began. The rock-wall, now removed, that flanked the continuation southward of the Narrow Gallery, traversed the hall; but, though standing, it was separated by an open crack from the roof, and had been undermined, as it stood on sand beneath. As we enter the Hywna Hall the pendent portions of partitions on the left show the same course of events. Portion of a deep, narrow partition between the two eastern galleries still stands, being propped by a block. ‘he pale sand that filled the deep trenches at each side of it was particularly rich in bones of Bear and Rein- deer. This hall was rich in animal remains. The masses of blocks at its northern end were covered with a lower stalagmite floor, resting in places on sand, but having fallen in where the latter was washed away. Under this stalagmite covering, and sometimes adhering to it beneath, were numerous bones and teeth. The part near the passage that leads to the Elephant Hall (5) had also a lower stalagmite tloor; and partly in this, partly in rubble, were several pieces of a Mammoth’'s skeleton, the long bones wanting their extremities, as though these had been gnawed away, and a Hyena’s jaw was found near them. lu the hollow of one of these large bones was a tooth of a young Mammoth. 1t was noticed that the bones found in the rubble were coated with hardened mud, and this was observed in many parts of the cave-system where bones were in rubble. This suggests that bones and limestone frag- ments bad been contained in a bed of sand that was drained away by waters that bore a muddy deposit which covered the bones. At the north-east corner of the Hyana Hall the irregular Sand Gallery, on a lower level, is plainly one of those water-courses that at a later time worked their way athwart the north and south system of galleries. 1t has quantities of loose, pale sand; but, with the exception of a few bits of ancient Scuarrr, Seymour, AND Newron—Custlepook Cave. 3T bones (probably drifted out of the higher sand-beds), it only contained fox- earth bones. Ifuge earthfalls, intrusions from roof-fissures, occupy some of the west and nearly the whole of the south side of the hall. In removing some of the latter we found dark, wet earth, lumps of charcoal, and teeth of horse and ox—all foreign to the cave. The passage (5 to 5A) that leads from the Hyena Hall to the Wlephant Ilall has the upper stalagmite sheet very close to the roof. It had also a lower stalagmite floor of later formation. The KLePHANT HALL (5A, 6A, 6,11), so called from the shoulder-blade of a Mammoth, the first bone of that animal found by us. ‘his hall is divided into three portions by walls or masses of rock that formerly divided separate galleries. The principal of these, which starts from the eastern division, is the Long Gallery, which extends some 120 feet, and is the finest in the cave. Another short, descending gallery leads from an alcove in the east wall of the hall, and after a turn, made by a cross-gallery, opens into the abyss. The eastern wall of the Elephant Hall is undermined, and seems unaccountably supported, as a large crack separates it from the roof, similar to what we noticed in the Hyena Hall. Running north from the above alcove is a very small gallery, in which we discovered a skull of Mammoth in sand and rubble that partially filled it, and in a cavity communicating with this was the hip-bone. Other bones were in the neighbouring lower stalagmite, which floored much of this and the next division of the hall, and rested on a bed of dark sand. ‘I'his sand contained broken masses of older stalagmite, with skulls of Bear and Reindeer. About a foot under it was buried another stalagmite floor, which, though broken, seemed to be im situ. Under this again was a paler sand. We encountered a huge mass of limestone covered with the former stalagmite that contained Mammoth’s bones. ‘Chis mass may have been the eastern wall of the Long Gallery, which had fallen down before the bones were deposited, as a rib of the Elephant was embedded on top of it. All that remains of the Long Gallery here, before we come to the fallen rock called the Tortoise (6), is the apex in the roof and the deep trench which is cut in a bed of the harder rock that dips to the north. A similar sloping rock-bed, with the trench excavated in it, was found in the next division of this hall, indicating a gallery whose tunnel-structure and sides are gone. Here the rock was buried deep under a bed of varied inter- stratified sands, which yielded no bones but those of Rabbit, Fox, and the prey of the latter. A rock-mass of fallen roof, 11 feet by 7 feet, lay over it. The vicinity of the north wall, however, produced Mammoth, Bear, and Reindeer, some bones being cemented to it. The western or third division of the Elephant Hall is encumbered with huge fallen masses of the roof, but 38 Proceedings of the Royal Irish Academy. in its north-east corner is a small gallery communicating with one of the orifices in the crags near the entrance. Returning to the Long Gallery (6, 7, 16) of which a fine view may be seen from the south end of the Tortoise block that spans it (see Plate VII, cross-sections), we find the tunnel very large, with pronounced shelf and deep trench, at the enlarged bottom of which it is perforated and crossed by the low-level water-courses that cut across the galleries. The principal of these descended into a cavity called the Lowest Cellar, which communicated with the bottom of the Abyss, and the Straddling Gallery east of it. ‘This great swallow-hole had undermined the eastern wall of the gallery, which fell, and its ruins now lie, coated with stalagmite, at the bottom of the Bride’s Hall, a portion of the Long Gallery whose roof is adorned with innumerable small stalactites. A diagonal gallery here crosses it, and from this crossing south- wards the Long Gallery possesses the longest stretch of upper stalagmite remaining unbroken in the cave. There is, however, a piece of it in the roof further south, and between the two the upper stalagmite has fallen, and its wreck is lodged in masses in the jaws of the trench, being cemented together by fresher stalagmite. In spite of being so splendid an example of a gallery, it contained but few fossil bones, its sand-bed having been almost all washed away. It ends in an earthfall, but has there an opening to the east into a chamber that has communication by a deep passage with the Abyss, which is parallel with the Long Gallery. FAIRY-LAND. The Abyss (9). We have crossed the line of this gallery (between 2 and 3) where we find it in the Little Narrow Gallery, and we have traced one side of it which remains in the southern part of the Hyzna Hall. We have found the southern extension of the Abyss running parallel with the Long Gallery in that direction; but it is the central portion, where it is best preserved, that now demands our attention. We enter it by a short passage descending from the alcove in the Elephant Hall, and there it is choked northwards by a high bed of sand surmounted by an earthfall and fallen blocks. From this point southward the Abyss is clear for 30 or 40 feet. Its great height and depth seem unbroken by tunnel or shelf. Portions of the upper stalagmite sheeting remain overhead, the west wall being perpendicular both above and below this stalagmite, and the gallery seems all one great trench which does not narrow nor expand much anywhere, though its dark sides are uneven and rough. A high talus of sand sloped down from where we now enter the Abyss, and this contained Scnarrr, SuyMour, AnD Newron— Castlepook Cave. 39 bones of Mammoth, &e., but we worked it out in 1904, piling the sand further south, It was covered with rubble, under which were 2 or 3 feet of red-brown bone-sand, then pale, barren sand, then again red-brown sand with bones, and lowest black, barren sand. Vertebrae and bits of tusk of Mammoth were in the dark red sand as well as in rubble and breccia, and the leg bone of a Bear was placed perpendicularly in the sand-bed, On the east side of the Abyss ‘about No. 7 level) is a series of openings into the next gallery that remind one of a clerestory, and the bases of this series run in an inclined plane of bedding, dipping north. They were apparently made by water which ran from this gallery into the next or vice versa, and in the orifice by which we squeeze into the next, the Straddling Gallery, Mr. lt. Evans found a Bear’s femur. At the lowest level explored are water channels of a much later period communicating with the galleries on both sides of the Abyss. The StRADDLING GALLERY (10) is dark, narrow, and uniform in breadth like the Abyss, but the upper stalagmite is continuous overhead. Entering through one of the clerestory openings (between 9 and 10) one finds ledges on each side which afford footing far above the bottom. Proceeding along these a hole is reached on the east side by which one passes into the third gallery, and a corresponding opening leads into the fourth gallery of this section, called Fairy-land. At the bottom of the Straddling Gallery was a bed of sand that contained bones of Bear and Reindeer. The Tuirp GALLERY. At the point where we enter the roof to the right shows a small twin gallery ; the partition must have fallen before the forma- tion of the upper stalagmite, as this extends across both it and the main gallery in an unbroken sheet. This Third Gallery contained in places three distinct floors of stalagmite at different levels, the lower floors being chiefly composed of broken masses united together, and such a secondary floor was found even under upper stalagmite unbroken. It was prolific in bones, chiefly below the third stalagmite floor, and these were coated with hardened mud. Though the curve of the roof widens above the upper stalagmite more than the perpendicular walls below, yet the tunnel shape is not developed either here or elsewhere in Fairy-land, and the deep trench, though represented, is not much narrower than the body of the gallery. There are side openings in the walls far above the level of the clerestory in the Abyss. Near the southern end of this long gallery is an opening low down on the west side that leads into deep, empty caverns and swallow-holes in the line of the Straddling Gallery. The FourrH Gauiery. The roof is duplicated as in the Third Gallery It shows the simple vaulting above the upper stalagmite, which remains in 40 Proceedings of the Royal Irish Academy. two places, the walls being approximately perpendicular down to the level of the trench, which is poorly developed. A secondary stalagmite floor was formed where the wreck of the upper had fallen on a sand-bed; under this was a hollow where the sand had been partially washed away, and there, in and under the second floor of stalagmite, were many bones, ¢.g. a leg bone of Mammoth and others, some of which adhered to the bottom of the stalagmite. We found also the remarkable feature (met with in the ‘Third Gallery) of the secondary floor of broken stalagmite extending under the unbroken upper stalagmite. The sand-bed stretched northward, and was rich in animal remains even where the roof became quite broken up and on a high level. There the walls are split and crushed by pressure, and an opening exists into the Fiera and StxtH GALLERIES, which are narrow, dark, and nearly empty, but a Mammoth’s bone was found in the fifth, and bits of bone in the sixth. From the central portion of the Elephant Hall (11) one creeps through a shallow passage into the Horse GALLERY (12a to 14). On entering a very striking view is obtained of its wide tunnel and upper stalagmite which forms a bridge or ceiling beneath the apex. Smaller galleries run on each side, one of which has a large cross-opening into the Horse Gallery. The sand in the latter was nearly barren of bones. Further on (towards 14) the whole side of the trench has slipped out into the centre, and this mass of rock we called the Horse. From (12) one of the few east and west passages of any size leads into the Bive or Farry HALt (1°), named from its broad, flattish ceiling of blue lime- stone with numberless small stalactites. ‘The western wall shows hollowing like the side of a tunnel. At the north end a wide tunnel gallery leads some distance until it is blocked up to the roof with pale sand which fills most of the hall to a depth of several feet. As the above gallery points towards orifices in Hurley’s part of the quarry outside, it seems that this profusion of pale, barren sand was drifted into the hall from the north as it was into the Goat House a little further east, and into the central part of the Elephant Hall. Several smaller galleries branch out of the Blue Hall northwards. We excavated one on the west side. ‘The shelf was found about 7 feet below the surface of the sand, and the trench in the rock was 12 or 14 inches wide. It contained some broken-up stalagmite, but this excavation yielded only one or two bones. In the centre of the hall beneath the great bed of pale sand we came on a broken-up stalagmite floor, and under that was a darker sand containing Bears’ and Reindeer’s bones. About the middle of the west side is an earthfall on top of which is an orifice that opens into a series of small galleries and swallow-holes. Further ScuarFr, Seymour, anp Newron—Castlepook Cave. 41 west again is a wider cavity having a floor of limestone rubble as though the sand had been washed out. From the southern extension of the Blue Hall a rounded opening worn in the rock leads into the Fat Man’s Passage, a deep, narrow gallery with regularly vaulted roof, which has another orifice on its east side opposite that by which we enter. ‘hese must have conveyed the water of the gallery which stops north of them. Squeezing sideways along the Fat Man’s Passage we come to ‘The HALL OF THE AGONIES (17). This with the HALL oF THE EARTHQUAKES (14-15) opening to the south of the Horse Gallery, are rugged open spaces encumbered with fallen masses of the roof. The gallery structure can only be traced where it issues from them. Continuing southward along what was a Double Gallery (17-18) we find the double torm at one part where we pass under a piece of the upper stalagmite. After this the gallery grows narrow with a deep sand-bed which, however, proved barren of animal relics. Before coming to the doorway of Hyena Land the gallery becomes vaulted, and one can stand upright; but at the Portal (18) the upper stalagmite begins, leaving so little space above our floor that one must kneel to enter, and here the passage was so blocked by limestone that there was barely room to pass the body through about 3 feet above the floor until we opened up a way. Some of the blocks are still incorporated with the stalagmite. It seems probable that the impediment here was such as to cut off Hyana LAnp, the section which we now enter. On working within this portal we at once unearthed such abundance of bones, including those of Hyzenas, as we had never yet found ; and this richness of animal remains continued through all portions of this very distinct section of the cave. It presented special features, the gallery structure being modified as follows :—The sand-bed was never far below the roof, seldom allowing one to stand upright; all the galleries that we worked had an abundance of it often packed tightly. The roof-vaulting, where present, consisted of a small pointed arch above the upper stalagmite, and the swelling tunnel was absent as well as the profoundly deep trench. Where the trench existed it was in a sloping rock-bed discon- nected with the roof in an open space chamber or hall. There was a series of east and west openings connecting the galleries and in some cases terminating them. There is moreover no steady air-current here, such as we used to find elsewhere, and at times candles burn dimly in Hyena Land. The nearness to the surface was realized when we heard a mowing-machine overhead. In R.I.A. PROC., VOL, XXXIV, SECT. B. [F] 42 Proceedings of the Royal Irish Academy the galleries of this section we found worn sandstone cobbles in the sand- beds. On passing through the portal and crossing transversely two low galleries filled up in our working (18-19) we came to two consisting of trenches in a bed of rock over which a low hall (19) extends westwards. These four galleries were called the Quadruple Set. A breach in the wall of the trenches between the third and fourth of these was filled with rubble, and in this was a radius of adult Mammoth coated with hardened mud. — This fourth gallery leads into the Threatening Gallery where the thick upper stalgamite holds on by one side only, supporting blocks of limestone. This gallery, rich in bones, had been inhabited by Hyznas, of which the most perfect skull that we found was only 8 inches below the surface of the sand. These animals appear to have made the Mammoth their prey especially in the very young state, from the remains we found here; and the presence of bones of Reindeer packed into this passage, and similar places, indicated that it co-existed with the Hyena. Remains of Bear were among the deepest that we found. A few worn pieces of sandstone occurred in the Threatening Gallery along with the limestone pieces. It terminates in a cul-de-sac, but close to this are openings right and left. That on the east communicates with the Aged Carnivores Gallery which here ends in an avalanche of stone. The orifice on the west side opens into a short piece of gallery called Dalton Gallery, where the globular head of a Mammoth’s femur had been left and was found loose on the surface coated with mud. Entering Dalton Gallery by a sloping water-worn orifice we pass on out of it by a corresponding one, both these openings being con- siderably above the bottoms of the galleries, Below the level of these side openings in Dalton Gallery was a bed of pale sand 15 inches deep, with dark sand under it, and a paler sand still beneath these. Bones were found in both these sands, most numerously in the pale one. Hyena remains were abundant also, a spine of a Mammoth’s vertebra, and ends of Reindeer’s bones, all less than 2 feet deep. Under this was a buried stalagmite floor, plainly that which had fallen from overhead. A broken skull of Bear was 1 foot below the surface, a piece of the broken stalagmite resting on it, and we continued to find Bears’ remains down to 3 feet deep. While excavating it was remarked that both in this and the last gallery the remains of Hyena were usually near the surface, while bones of Bear were among the deepest. Beyond Dalton Gallery we found a barren stony chamber, and we return through the Threatening Gallery to the low hall (now filled up). This extends westwards over benches of rock that dip to the north. On this rocky floor sand and bones had accumulated, and a thick stalagmite floor had formed. We took bones of Reindeer out of this floor and beneath it. It is evident that the gallery-structure had here disappeared, and that sand had been Scuarrr, Seymour, anp Newron—Castlepook Cave. 43 deposited on the rocky bottom, and the hall thus formed inhabited by the extinct animals, and also that stalagmite had formed before that fauna had disappeared. One difference between this and the former halls is that the space between the upper stalagmite (observable in parts of galleries that remain) and this second or lower floor is very much less. We had to cut bones out of this stalagmite with a cold chisel. Others of Mammoth, Bear, Hyena, and Reindeer were met with either loose among rubble or in sand under this stalagmite; a tibia and horns of Reindeer showed marks of the teeth of rodents, but had been subsequently coated with sandy mud over the tooth-marks, showing that these were ancient. The west side of the low hall was open into the side of the largest gallery in Hyena Land. It extends north and south beyond the limits of the low hall, and from the teeth of old Bears and Hyznas found there IJ called it the Gallery of the Aged Carnivores (19). Portions of the upper stalagmite remain two or three inches below the roof bridging over the gallery, seven feet wide. Under this was an empty space, and then pale, barren sand on the surface covering the darker sand-bed, which was rich in bones. Among the first of these was part of a Mammoth’s mandible with molar tooth. We first excavated the sand northward to a depth of 3 feet until at 19 feet we were stopped by an earthfall, and we then worked back southwards, digging deeper. The bone-sand contained great numbers of remains of the above animals from Mammoth to Lemming, including two bones of Irish Elk, and notably jaws and teeth of Hyzena, a cranium of which was found deep in a fissure or swallow-hole on the east side. There were also coprolites of this animal, and remains of very old and very young Bears. Reindeers’ bones were plentiful, and the recesses at the sides of the gailery were specially prolific. We then worked this gallery southward (where it was 8 feet wide) for 17 feet, and there a pile of fallen stone and earth stopped further progress. In this part we discovered a broken skull and other bones of Wolf, and a very brittle skull of Fox was 3 feet deep under sand packed with stones, At various points and depths were rounded pieces of sandstone or cobbles. These could not have been intruded with earthfalls at a later date unless the whole sand-bed with its profusion of bones was vemanié. The objection to such a supposition is that delicate skulls lay in the same sand-stratum with heavy blocks of limestone and stalagmite, and where the latter had fallen its remains lay buried 3 feet or more under the bones. The pale barren sand had indeed been drifted in on the surface, but beneath that all seemed undisturbed since the deposition of the animal remains. On the west side of the above gallery is a large water-worn orifice, sloping north, which leads into the parallel GALLERY oF THE ELEPHANTS’ TEETH LF 2] 44 Proceedings of the Royal Irish Academy. This, at its southern end, has the vaulted roof, the upper stalagmite having there fallen, and its wrecks were found under six feet of sand. Where the gallery of the Irish Elk branches off, opposite the above orifice, a very deep bed of sand occurred, which contained a surprising assemblage of young and old Reindeers’ remains, down to the low level of nine feet below the upper stalagmite. Near them had been a swallow-hole, and it seemed as if bodies of Reindeer had been drawn down by the vortex of water to this vicinity. The angular space between the two galleries (of the Elephants’ Teeth and of the Irish Elk) had a horizontal vacant area, at a high level, in which stands the upper stalagmite sheet, crossing both these galleries as well, within a few inches of the irregular roof. It is found by Professor Seymour to be practically on the same level as the upper stalagmite in the distant parts of the cave, and marks the horizon of its earliest sand-bed of which we have evidence. On this bed, now gone, the stalagmite must have been formed. As we worked northward we found teeth and bones of Mammoth, Hyena, and Wolf in the sand, which closed up to the upper stalagmite, and after that an earthfall stopped our progess. In this gallery, as well as in that of the Aged Carnivores, was a bed of pale, barren sand, about 18 inches deep above the bone-sand, and this deepened at the swallow-hole above men- tioned. The GALLERY OF THE IRISH ELK has an irregular, slightly-vaulted roof over the upper stalagmite (which remains for 13 feet), and perpendicular sides until we dug down, where we found they closed in V form. At six to seven feet from the corner of the last gallery we found a large limestone block from two feet to three feet below the surface of the sand, with bones of Hyzena and Mammoth. If this block fell at any time from the roof it must have done so before the sheet of upper stalagmite (which is still above) was formed. This would tend to show that the sand and bones among which it lay were more ancient than the upper stalagmite; but in the last two galleries the bone- sand lay over the fallen fragments of the upper stalagmite. Sandstone cobbles were found between the pale upper sand and the coarse, dark bone- sand. ‘l'here were various bones of Hyena, Mammoth, Reindeer, and Irish Elk. Of the latter we got a piece of the beam of one antler, apparently gnawed and discarded. The irregular D1aGoNAL GALLERY or chamber on the left or south yielded more of the above bones. At 18 feet the Gallery of the Irish Elk is intersected by a narrow north- and-south passage, just where its direct course is barred by an earthfall (20). This we called the GALLERY OF THE VERTEBRA, from a cervical vertebra of a Scuarrr, SeyMour, anpd Newron—Castlepook Cave. 45 Mammoth found in it, besides which, on digging deep, we got bones and teeth of the usual fossil mammals. At its southern end this opened on its east side into the Brar’s Drn, which extended some 20 feet further south, and which terminates its direct course abruptly at each end. At either end it has openings east and west. That opposite the end of the Vertebra Gallery leads into an extension of the gallery of the Swallow-hole (below), and it proved to be very deep and full of bone-sand, in which we got a half mandible of adult Hyena, with Reindeer bones both above and beneath it, as far down as six feet below the surface; also remains of Bear and Mammoth. The Brar’s Den has a simple vaulted roof, with perpendicular sides, which, however, narrowed beneath the sand-bed, leaving barely room to work in. This sand, which was muddy or earthy, contained blocks and rubble that increased as we descended. We dug in places four feet deep, and the first foot or so contained hardly any bones; but the bed below that was the richest I have seen, so that we filled two riddles with remains of Mammoth, Bear, and Reindeer. The richness of these deep, narrow galleries is due to the bones having been packed into close limits, and buried deep, where they lay undisturbed. At its southern end the Bear’s Den led eastward by another sloping opening into a chamber that crossed the trenches of two narrow galleries. The nearest or western one contained a swallow-hole, in which was found a fine Mammoth’s tooth, loose, under a stone. Both these galleries, and the chamber that contained them, yielded from their sand-beds a number of bones. On the opposite or west side of the Bear’s Den (21) the sloping orifice led into the GALLERY OF THE HuUMERUS, so named from a Mammoth’s humerus discovered in a deep cavity, packed in with pieces of limestone, under a cake of brecciated sand. The extremities of this huge bone had been much gnawed. The southern extremities of this, and of the Gallery of the Swallow-hole, were too narrow to pursue. An examination of the Low Hall (19-21) shows that its structure is analogous to that of the smaller chamber of the Swallow-hole Gallery. In each of these a wide vacancy was formed, above the bedding-plane, of a harder stratum of rock, leaving the deep trenches of the galleries in the latter. In the Low Hall this plane slopes up above the level of the upper stalagmite, at the mouth of the Threatening Gallery, where it depends from one side, laden with blocks. A similar slope may be seen in the Elephant Hall, where the trench of the Long Gallery commenced. ‘lhe dip in these cases is always to the north, and the same slope is observable in the series of east and west openings at 46 Proceedings of the Royal Irish Academy. the southern end of the Threatening Gallery, in the opening leading into that of the Elephants’ Teeth, and at each end of the Bear’s Den, left and right. Such openings are characteristic of this section (Hyzna Land), and plainly conveyed the currents of streams that formed the galleries at a much earlier period than those swallow-holes and cross orifices found at the bottoms of deep trenches (as in the Long Gallery and the Abyss), which point to a late chapter in the cave’s history, and which drained away the sand-bed in their vieinity.” 2.—GEOLOGY. The remarks which follow are based on a personal examination of the cave and surrounding district made on the occasion of several visits to Castlepook in company with Mr. Ussher, whose loss, im common with all who had the pleasure of his acquaintance, I most deeply deplore. Only those, perhaps, whe day after day watched his ceaseless energy in the cave and the unvarying enthusiasm with which he worked, in spite of conditions trying even to men of half his age, can appreciate what an irreparable loss that particular branch of research in this country has sustained by his death. On several occasions it was my privilege to help him by laying out the main lines of the survey on which his beautiful plan of the cave was based. Those who have attempted with an ordinary field theodolite to carry out a cave survey can realize the severe physical discomforts involved in the process; and not the smallest share of these was borne by Mr. Ussher, and in a manner that compelled admiration. On my last visit, only a few weeks before his death, I spent three days running a series of levels round the whole of the cave, in order to determine the horizon of the shelf and stalagmite floor, &c., and again he gave me most material assistance, and took copious notes from which numerous cross-sections were drawn, some being here reproduced (see map). One result of the levelling was a source of great satisfaction to him as confirming a pre-conceived opinion that the main stalagmite floor developed to a greater or less degree in the various passages in the cave was all on one horizon, and belonged to the same period of formation. The cave is formed in whet is now a bluff or knoll of limestone of Carboniferous Age, the Lower Limestone of the Geological Survey. A few miles to the north are the Ballyhoura Mountains of Old Red Sandstone Age, rising to a height of some 1400 feet. As is usually the case, the cave consists of a number of passages formed by enlargement of the vertical cross-joints of the rock. Those of one parallel Scuarrr, Seymour, anv Newron—Castlepook Cave. 47 set of passages running approximately 10° east of north are more numerous, extensive, and regular, the cross-connecting series being fewer in number, of more irregular form, and generally much smaller, sometimes mere tortuous tunnel-shaped tubes through which one can pass only in a prone position, The main set has generally a characteristic shape in cross-section, which may be likened to a mushroom, consisting of two parts (see map; fig. 3), an upper helmet-shaped portion and a lower narrow and deep trench (the stem), a nearly horizontal shelf occurring at the junction. In a few instances a miniature replica is seen capping the principal cavity. From a study of the form of the passages, it becomes evident that they were initiated by an underground drainage system which produced small tunnels, the bottoms of which coincided with a main bedding plane of the limestone (see especially fig. 4 on the map). A comparatively dry period followed, during which the stalagmite floor was accumulated on top of a sandy deposit containing numerous bones, belonging to various animal types, A wet period supervened, which allowed of the enlargement of the old passages, but on a lower level—down in fact to the next bedding plane, possibly lower. In some cases the previously formed stalagmite floor has been thus undermined, and has fallen down and become lodged in lower levels, and the bones contained in it thereby become associated with bones of a more recent period. There are some traces also of another stalagmite floor at a still lower level, and apparently in situ. The highest stalagmite floor is, of course, the oldest, and those below are (mostly reconstructed) more and more recent, so that there is an inverted stratigraphical sequence as regards the fossils. The relative width of the upper and the lower portions of the typical passage is capable of more than one explanation. Thus it might be the result of differential erosion on two beds of limestone of very different solubility. The bed, however, in which the narrow lower portion is mainly cut is occasionally excavated to the same width as the upper portion. See fig. 2, for instance. I am inclined therefore to attribute the formation of the main tunnels between the principal stalagmite floor and the conspicuous shelf some three feet below, to the maintenance for a sufficient time of the outflow level or discharge level of the water in circulation, at the level of the shelf. Under such circumstances, any water below that level would be practically stagnant, and would be quickly saturated, and thus incapable of further solution. The upper and lower portions of the passages would, therefore, be forming at the same period, but only in the upper part would considerable solution be taking place. The lowering of the discharge-level would initiate solution at 48 Proceedings of the Royal Irish Academy. a lower level, the amount varying locally even on the same level according to the volume of water circulating in the passage. It also appears to me probable that there would be a tendency for the postulated reduced level of discharge to maintain itself for some considerable time at the level of the next lower bedding plane in the limestone, and thus allow of considerable solution at that level, causing undercutting and collapse of dividing walls, which is a feature of the Castlepook cave, and a process which is indeed taking place even now on a limited scale: vide “ Threatening Gallery.” Referring to the map of the cave (Plate VII), most of the points just touched on will be apparent. The principal shelf is a conspicuous feature, forming a narrow border to the deep clefts or chasms which were once filled with sand and gravel (indicated by dots), all of which was removed by Mr. Ussher. It was by such excavation that the cave was opened up, the map showing the present extent of the openings which are accessible, but by no means indicating the real extent of this cave, which is capable of considerably more excavation and enlargement at the hands of future workers. The geological evidence as to the age of the cave is of an inconclusive character, and may be summarized as follows:—There is a high degree of probability that the whole of the region, including the highest summits in the vicinity of Castlepook, was overridden by the great ice sheet of the Glacial Epoch at the period of its maximum development. The resulting boulder-clay, while consisting largely of local material, contains a very few pebbles of a granite similar to that got east and south of Limerick, and most probably derived from a source in Connemara. Some boulders of this rock occur in the walls of the fields on the slopes above the caves, and Mr. G. W. Lamplugh as well as the writer got them in the stream beds in the vicinity. No foreign erratic of any description was found amongst the hundreds of stones taken from the cave by Mr. Ussher and examined by the writer. Many boxes of specimens were also seen at the Museum, but all were of local rocks only. This (confessedly negative) evidence would seem to indicate that there was no post-glacial, late, or even perhaps inter-glacial circulation of water in the cave so far as our evidence goes, otherwise erratics of granite would be likely to occur in the deposits. Their apparent absence can be readily accounted for on the supposition that the pre-glacial orifices were sealed up under a covering of boulder-clay. Hence it would appear likely that the cave was excavated by a pre-glacial drainage system—an inference which is strongly supported by the largely pre-glacial facies of the fauna associated with the cave deposits. Between this fauna and the fox-earth with remains of domestic animals Scuarrr, Seymour, and Newron Custlepook Cave. 49 there is an obvious hiatus, which no doubt represents a time-interval com- mencing with the sealing-up of the cave orifices on the initiation of the Glacial Period and ending when denudation and quarrying operations exposed them again to view from beneath the mantle of lower boulder-clay. The only argument for a post-glacial age for the cave les in the fact that the drainage levels of the latter and the level of the present drainage system of the district are not more than about 20 feet apart vertically, possibly even less. If the cave is of pre-glacial age, as seems most probable to the writer, this implies an exceptionally slow rate of denudation locally during the time which has elapsed since the end of the Glacial Period. This, however, may be accounted for on the hypothesis that the present drainage system was established at a period long after the disappearance of the ice sheet, and may possibly have been preceded by a system of drainage in some other direction. iver diversion of a temporary or even permanent nature has been a common result of the change of local topography by accumulations of boulder-clay following the Ice Age. The palaeontological evidence, however, in addition to the geological, seems to the writer to point to the cave having been formed before the Glacial Period. The geological evidence is therefore not conclusive one way or another, though inclining, on the apparent absence of any erratic in the cave deposits, to the view that the cave is of pre- glacial age. 3.—ANIMAL REMAINS (except Birds). Natrerer’s Bar (JZyotis natterert). Only two long bones of bats were found in the cave, one of which belonged to this species, the other to the next. Both looked modern and are probably fairly recent. According to the late Major Barrett-Hamilton,’ this species still lives in Co, Cork. - SMALL Horsk-SHoe Bat (Rhinolophus hipposideros). This bat was found inhabiting the caves of Co. Clare, and its bones were there met with in the upper and the lower stratum. In Castlepook Cave no recent specimens were observed, only a single long bone which was among the remains of Reindeer, Irish Elk, and Hyzena. ! Barrett-Hamilton, G. H. H.: ‘‘ British Mammals,” pt. iv, p. 180, London. R.I.A. PROC,, VOL. XXXIV, SECT. B, [@] 50 Proceedings of the Royal Irish Academy. Domestic Cat (Felis domestica), The cave deposits contained but few remains of cats, and most of these were in the first stratum, together with rabbit, horse, sheep, and ox. About half of the teeth and bones of the cats found belonged to adult animals, the remainder being young. All the cats were undersized and below the average stature. The lower carnassial tooth seemed somewhat larger than is usually the case in domestic cats. Nevertheless, all of these cat-remains must be referred to small domestic cats, for none of them showed signs of antiquity. No certain evidence in this cave of the Wild Cat could be detected—an animal which had been observed in the Clare caves and in Ballinamintra Cave. Irish Stoat (Mustela hibernica, Thomas and Barrett-Hamilton) = Putorius ermineus hibernicus. During the three years that Mr. Ussher was actively engaged in the Castlepook Cave excavations not a single tooth or bone of any Stoat-like creature was discovered. When he resumed operations shortly before his death in the summer of 1913, he met with the left ramus of a small jaw, which proved to be that of an Irish Stoat. As it was found under a layer of stalagmite together with the bones of Mammoth, Bear, and Reindeer, there can be no doubt of its antiquity. Irish Stoat bones have occurred in the Caves of Kesh and those of Co. Clare. In the former, the bones were associated with those of Lemming and Field Mouse. These facts indicate that the Irish Stoat is one of the early inhabitants of Ireland, and that it came with animals long since extinct in this country. Fox (Vulpes vulpes = Canis vulpes). Fox remains occur abundantly in all or most of the Irish Caves that have been critically examined, and the Fox is still plentiful in Ireland at the present day. In my report on the Fox of the Clare Caves' I dwelt on the difficulty of discriminating between small forms of the Common Fox and the Arctic Fox. In the present case, some of the smaller teeth and bones were recent-looking, and thus probably belonged to the Common Fox. Others were distinctly ancient in appearance, yet, although small, they seemed to be referable to this species. » “> CS) eer : A a y q >, SS Y, YY a) zupds A TU yy Dy gu ices “S12 bip ay > = SYS ei Hy Ly RSA yy Y CY ND) ne mn) y Duns me) mM F( aR ‘S Y l i) ~ — ‘TIX FLYVTd "q “LOUS “AIXXX “IOA “AVOY “Tw ‘90uT Proc. R. I. Acab., VoL. XXXIV, SEcr. B. 22 CARPENTER AND POLLARD—LATERAL SPIRACLES AND TRACHEAL TUBES O1 I PLATE XIII. THE LARVA OF H PC DERMA. eh paras £ . 1 ns ied “J Aji v eos?) V. SYNTHESES OF SOME NEW SUBSTANTIVE DYES DERIVED FROM BENZIDINE-SULPHONE. By HUGH RYAN, D.Sc.; JOSEPH ALGAR, D.Sc.; anp PHILIP O'CONNELL, M.Sc., University College, Dublin. Read Aprit 22. Published Avousr 13, 1918. SinckE the compounds obtained by coupling naphthols and amines with the tetrazo derivatives of benzidine are amongst the most important of the direct cotton dyes, any further investigation of substances of this type is likely to prove of interest and importance. The present paper deals with a number of dyes of the benzidine type, obtained by coupling hydroxy and amino compounds with the tetrazo derivative of benzidine-sulphone-disulphonic acid. The preparation of benzidine-sulphone-disulphonic acid is described by P. Griess and C, Duisberg (Ber, 22, p. 2459), who state that it combines, when tetrazotised, with naphthols and naphthol-sulphonic acids, yielding purple to violet dyes, and with naphthylamines and naphthylamine-sulphonie acids, forming red or bluish-violet colouring matters. References also occur in Patent literature (Friedlander I, p. 495, and II, p. 408, and Baeyer & Co., D. R. P. 27,954 and 51,497) to red to blue dyes obtained by coupling the above-mentioned tetrazo compound with a- and /3-naphthylamines, and with Schiffer’s, “R” and “G” acids. None of the dyes mentioned, however, seem to have been isolated in a pure condition, nor does it appear that any effort was made to determine their constitution and chemical properties. The various colouring matters described in this communication were obtained in a pure condition as sodium salts. Repeated efforts to obtain them as crystalline substances were unsuccessful. ‘The coupling of the different hydroxy and amino compounds with the tetrazo derivative took place readily in somewhat alkaline solution, and, as a rule, the coupling was complete after R.I.A. PROC., VOL. XXXIV,, SECT. B. (1) 86 Proceedings of the Royal Trish Academy. about twenty-four hours. The yield of crude product obtained was usually almost quantitative. All the compounds mentioned acted as direct dyes towards cotton, and the colours obtained were unaffected by washing. The dyeing was effected by dissolving “15 gram of the dye stuff (3 per cent. on the weight of cotton) in a little warm water, and adding it to 100 cc. of water containing 2 per cent. of sodium chloride and 2 per cent. of sodium carbonate. To the dyé-bath was added 5 grams of the thoroughly washed cotton, and the whole was slowly raised to boiling point, and finally boiled for one hour. The colours obtained with the different compounds varied over a wide range, as is shown in the following table -— AbsuNcT. DescriPrion. Cotoun ix Soturion.| Suave on Cotton. Sulphanilic acid, Orange-red powder, Deep red, Canary yellow. _ Dimethylaniline, Dark blue powder, Deep red, .. Deep purple. Naphthionic acid, Dark blue powder, Bright red, Purple. a-Naphthylamine, .. | Dark red powder, -- | Deepred, .. +. | Navy blue. 8-Naphthylamine, .. | Reddish powder, Deep red, .. Violet red. Salicylic acid, Reddish brown powder, .. | Yellow, Orange. Schaffer’s acid, Dark blue powder, Deep red, Light purple. “«R-acid,”” ., Reddish blue powder, Deep red, Violet red. “¢G-acid,” .. Reddish powder, .. Deep red, Pink. ‘“‘H-acid,” .. Blue powder, Deep blue, .. Light blue. Pyrocatechin, Dark blue powder, Dark golden, Light brown. Resorcin, Dark purple powder, Light cherry-red, .. | Maroon. Hydroquinone, Brown powder, Light amber, Buff. Pyrogallol, Chocolate brown powder, Light amber, Buff. Gallic acid, .. Dark brown powder, Light brown, Very light brown. It will be seen from the above table that the compounds obtained by coupling the tetrazo derivative with hydroquinone, pyrogallol, and gallic acid gave rather poor shades on cotton, whereas most of the other compounds gave good colours. There is a very striking difference in the shades produced when a- and (3-naphthylamines are used as adjuncts, the former giving a navy blue and the latter a violet-red colour. On the whole, the amino compounds seem to give bluer colours than the hydroxy compounds. Apart from the dyeing properties of these tetrazo derivatives, it is possible that they may prove of importance in the chemotherapy of Spirilla. Ryan, AuGar, and O’ConneL~—Syntheses of Some New Dyes. 87 The chemotherapeutic properties of the following dyes of the benzidine series have been examined :— Naga-red (Benzidine coupled with -naphthylamine-3-6-disulphonie acid). Trypan-red (Benzidine-mono-sulphonic acid coupled with [- naphthylamine-3-6-disulphonic acid). Trypan-blue (Tolidine coupled with amido-H-salt), Trypan-red was tested by Vassel (Comptes Rendus de la Soe. de Biol., 62 (1907), p. 414), and recognized as effective in the treatment of diseases due to Spirilla; and Manteufel (Arbeiten aus dem Kaiserlichen Gesundheitsamt, 29, No. 2) also recognized the beneficial effect of benzidine dyes. Mesnil’s trypan- violet-has also been employed in addition to his trypan-blue. An investigation of the chemotherapeutic properties of the dyes mentioned in this paper is at present in progress. EXPERIMENTAL Part. Tetrazotisation of Benzidine-Sulphone-Disulphonic Acid. A solution of 20°3 grams of benzidine-sulphone-disulphonie acid in 150 c.cs. of water containing 20 c.cs. of concentrated hydrochloric acid was cooled with ice to 5°C. 15 c.es. of concentrated hydrochloric acid were added, and the solution tetrazotised by-slowly adding, a few drops at a time, a solution of 7-2 grams of sodium nitrite in a little water. The tetrazo solution was stirred vigorously and the nitrite added until it gave a slight reaction with starch-iodide paper. 1. Coupling of Benzidine-Sulphone-Disulphonic Acid with Naphthionic Acid. Re SO3Na SO3sNa NH = = | (Y Sone —S0.Na \ 80,7 SO,Na—< Bs Benzidine-sulphone-disulphonie acid (20°3 grams) was dissolved in 150 c.cs. of water and tetrazotised. The tetrazo solution was slowly added to the “G-salt” solution, which was prepared by dissolving 40 grams of “G-salt” in 400 c.cs. of water, adding 10 grams of caustic soda, with 5 grams of sodium carbonate, and cooling to 15°C. During the addition the mixture was stirred mechanically, and the stirring was continued for five hours afterwards. After standing overnight, the mixture was heated to 80°C., and the colour pre- cipitated by saturation with common salt. The colouring matter was filtered and dried. The compound was purified by heating with dilute alcohol and filtering, when it separated out on cooling. It is a reddish amorphous powder, soluble in water to a deep-red solution, and acts on cotton as a direct dye, giving a pink shade. "2420 substance gave 10°3 c.cs. moist nitrogen at 16°C. and 768 mm. P., corresponding to N = 5:0, CH,,N,0.8;Na, requires N = 4°79. Ryan, Atcar, anv O’Conneti— Syntheses of Some New Dyes. 91 7. Coupling of Benzidine-Sulphone-Disulphonie Acid with “ H-Acid” (1 Amino-8-Naphthol-3-6-Disulphonice Acid). NHz OH SO3;Na SO3Na OH NH, eS) [les eh | (NOES =NaY Sa SEN ERE NN al SIE ONED RN | SOsNa_\ A _7—803Na Ne 56.7 SOsNa— A —S80sNa The tetrazotised solution of benzidine-sulphone-disulphonic acid was added slowly, with constant stirring, to a solution of “ H-acid,” which was prepared by dissolving 40 grams of “H-acid” in 450 c.es. of water, and adding 40 grams of sodium carbonate, the mixture being cooled to 15°C. When all was added, the mixture showed an alkaline reaction. The stirring was maintaimed vigorously for five hours afterwards, when the liquid was heated to 80° C., saturated with common salt, and filtered. It is a deep blue amorphous powder, giving.a deep blue colour in alkali solution and red in acid solution. It was purified by separation on cooling from hot dilute alcohol. The substance dyes cotton directly a light blue. ‘1880 substance gave 10°8 c.cs. moist nitrogen at 16°C. and 760 mm. P., corresponding to N = 6°67, Cy. HysN.O22.Na,S; requires N = 7:01. 8. Coupling of Benzidine-Sulphone-Disulphonic Acid with Schaffers Acid (3-Naphthol-6-Monosulphonic Acid). OH ea SO,Na OH I ih a (i TRENT AE NRE TNE EIEN OU oe ia oe ee nN Ny, Pw any? Van yy, Nog SO3Na SOsNa A solution of Schiiffers acid was prepared by dissolving 27 grams of the acid in 200 ccs. of water containing 35 grams of sodium carbonate, and cooling the mixture to 50°C. The mixture was stirred vigorously by means of a turbine, and the tetrazotised solution from 20°3 grams of benzidine- sulphone-disulphonie acid was slowly added; the stirring being maintained for five hours afterwards. On standing overnight the colouring matter was heated to 80°C., and precipitated by the addition of common salt. It was then filtered and dried. The compound was purified by heating with water and a little alcohol, then filtering and allowing to cool when the dye stuff 92 Proceedings of the Royul Irish Academy. separated out. This process was repeated a number of times in order to obtain the compound pure for analysis. 1940 substance gave 9°4 c.cs. moist nitrogen at 16° C. and 762mm. P., corresponding to N = 5°66, C,-H,.01S;Na,N, requires N — 5:80. The pure compound is a dark-bluish amorphous powder which gives a deep-red solution in water, and dyes unmordanted cotton a light-purple shade. 9. Coupling of Benzidine-Sulphone-Disulphonic Acid with Pyrocatechin. OH SOsNa SO.Na OH l a | l of 9 ee HO 8 a Foon ae Bal \ 50.7 \4 The benzidine-sulphone-disulphonie acid (22°3 grams) was tetrazotised in the usual manner, after which it was added to a cold solution of 30 grams of sodium carbonate with constant stirring. 12 grams of Pyrocatechin, dissolved in a little water, were added immediately, and the stirring continued foraday. After standing overnight the solution was heated on a water-bath to 80°C., and the colouring matter precipitated by the addition of common salt, when it was filtered at the pump, dried, and ground. It was purified as before by repeated separation, on cooling, from hot dilute alcohol. 1586 substance gave 10°8 c.es. moist nitrogen at 16°C. and 763 mm. P. corresponding to N = 7:90, C.,.H,,0,.N,Na.S, requires N = 8:10. The pure substance is a dark-bluish amorphous powder soluble in water to a dark-brown solution, and dyed cotton directly a light-brown. 10. Coupling of Benzidine-Sulphone-Disulphonic Acid with Resorcin. OH SO ,Na SO,Na OH |)\ SD OSS RL QE: 2 = Wake Se aor” pre The method adopted was similar to that used for Pyrocatechin. 22-3 grams of benzidine-sulphone-disulphonic acid were dissolved in 150 c.cs. of water, and tetrazotised with a concentrated solution of 7-2 grams of sodium Ryan, Agar, and O’Connett—Syntheses of Some New Dyes, 93 nitrite. This tetrazotised solution was added to a cold solution of 30 grams of sodium carbonate, the mixture being stirred vigorously during the addition. A solution of 12 grams of resorcin, in a little water, was added immediately, and the stirring continued for about 8 hours, when the mixture was allowed to stand overnight. It was then heated to 80°C., and the colouring matter precipitated by the addition of common salt, when it was filtered and dried. It was repeatedly purified from dilute alcohol. 2374 substance gave 15:6 c.cs. moist nitrogen at 16°C. and 765 mm. P., corresponding to N = 7:80, C.4H,,0;,.NiNa.8, requires N = 8°10. The pure compound is a dark-bluish amorphous powder, dissolving in water to a light cherry-red solution, and dyes cotton directly a beautiful maroon shade. 11. Coupling of Benzidine-Sulphone-Disulphonie Acid with Hydroquinone. OH SO3Na SO3Na OH RN ag Na aR aN (Ce ee Ne ares a ae Ne \s0.7 XG OH OH The benzidine-sulphone-disulphonice acid (22°35 grains) was dissolved in 150 c.cs. of water, and tetrazotised in the usual manner with 7-2 grams of sodium nitrite. The tetrazotised solution was added to a cold solution of 30 grams of sodium carbonate, the mixture being stirred mechanically, and 12 grams of hydroquinone, dissolved in a small quantity of water, was added immediately. After stirring the mixture vigorously for eight hours, it was left standing overnight. It was then heated on a water-bath to 80°C., and, having been maintained at this temperature for half an hour, common salt was added when the precipitated colouring matter was filtered and dried. The dye-stuff is a brown sandy-coloured amorphous powder giving a light amber-coloured solution in water. It was purified by separation on cooling from hot dilute alcohol, and acted on cotton as a direct dye giving a buff shade. ‘1850 substance gave 12°3c.es. moist nitrogen at 15° C. and 768 mm. P., corresponding to N = 7:96, C2sH,,0;.SsNa.N, requires N = 8:10. R.I.A. PROC., VOL. XXXIV, SECT. B. LN] 94 Proceedings of the Royal Irish Academy. 12. Coupling of Benzidine-Sulphone-Disulphonic Acid with Pyrogaltol. OH OH SO3Na S03Na | | u iia l oe ie, HO >—N = nS Son =n—< Son NA lt The method adopted was similar to that for the previous dihydroxyl compounds. The benzidine-sulphone-disulphonic acid having been tetrazotised and added with constant stirring to a cold solution of 30 grams of sodium carbonate, the solution of pyrogallol (21 grams) was added immediately. During the addition the mixture was stirred vigorously, the stirring being continued for eight hours afterwards. On standing overnight, the mixture was heated to 80°C., and saturated with common salt, when the colouring matter was precipitated. It was filtered at the pump, and purified for analysis by repeated separations on cooling from hot dilute alcohol. The purified compound is a chocolate-brown amorphous powder, giving a light amber solution in water. It dyes cotton directly a buff shade. 1030 substance gave 6°7 c.cs. moist nitrogen at 15°C. and 767 mm. P., corresponding to N = 7: C.,H,,0,,.N.Na,S, requires N = 7 13. Coupling of Benzidine-Sulphone-Disulphonic Acid with Gallic Acid. COONa SOsNa SO;Na COONa | | | POR ag gg IE OTE oe oO’ BO ee a Se eS HO—\_/—0H N\ s027 HO-\ /—0H OH aid OH Gallic acid (21 grams) was dissolved in 100 ¢.cs. of water, and the solution obtained by tetrazotising 20°3 grams of benzidine-sulphone-disulphonic acid was added with constant stirring, which was maintained mechanically for two days. After the mixture had been stirred for half an hour a solution of 35 grams of sodium carbonate was added, a few drops at a time, so that the whole was used up by the end of the second day. After standing overnight the mixture was heated to 80° C. and saturated with common salt when the dye-stuff was precipitated. It was purified by repeated separations, on cooling from hot dilute alcohol, and formed a dark-brownish amorphous powder soluble in water to a light-brown solution. The colouring matter dyes cotton directly a very light-brown. Ryan, ALGaR, AND O’ConnELL—Syntheses of Some New Dyes. 95 “2774 substance gave 15:2 ¢.cs. moist nitrogen at 16°C. and 761 mm. P. corresponding to N = 6°35, C.,H,,0;,NaiNiS; requires N = 6°54. 14, Coupling of Benzidine-Sulphone-Disulphonie Acid with Sulphanilic Acid. i SOsNa SO3Na NH: | YN BNO NN CS ee ea Y \s0.” ( SO3Na E SO3Na The tetrazotised solution of the benzidine-sulphone-disulphonic acid (20°3 grams) was poured into a cold concentrated solution of 58 grams of sulphanilic acid to which 35 grams of sodium carbonate had been added, and the mixture stirred mechanically for two days. After the first half-hour’s stirring a solution of 20 grams of sodium carbonate was slowly added so that the whole was used up by the end of the second day. On standing overnight the mixture was heated on a water-bath to 80°C., and the colouring matter precipitated by the addition of common salt, when it was filtered and dried. The compound was purified by bringing it quickly into solution in hot water, in which it was easily soluble to a deep-red solution, which, on cooling, deposited the dye-stuff as a deep orange-red amorphous powder. It was hardly soluble in cold water. In spite of successive purifications the substance was not obtained in a crystalline condition. The compound dyes cotton directly a canary-yellow shade. ‘1640 substance gave 13°5 c.cs. moist nitrogen at 17°C. and 770 mm. P., corresponding to N = 9°68, Cx1H,,OuN,NaS; requires N = 9°74. 15. Coupling of Benzidine-Sulphone-Disulphonic Acid with Dimethylaniline. SOsNa SO3Na CH; — = =| — CHs Nea i Oa NON NZ eNO ane Nos os NT Nee \N 802.7 The dimethylaniline (87 grams) was cooled in ice in a large beaker, and into it was poured, with constant stirring, the solution obtained by tetrazotising, in the usual manner, 20°3 grams of benzidine-sulphone- disulphonie acid. The mixture was stirred vigorously for two days. After stirring for half an hour a solution of 85 grams of sodium carbonate was 96 Proceedings of the Royal Irish Academy. slowly added, drop by drop, at such a rate that the total quantity was used up by the end of the second day. The mixture on standing overnight was heated on a water-bath to 80° C., saturated with common salt, and maintained at this temperature for half an hour. The precipitated colouring matter was filtered and purified by repeated separations, on cooling, from hot dilute alcohol. ‘1624 substance gave 16-2 c.cs, moist nitrogen at 16°C. and 764 mm. P., corresponding to N = 11°68, C.,H.,O,N,Na.S; requires N = 11°76. The dye-stuff is a dark-bluish amorphous powder soluble in water to a deep-red solution. It gives direct on cotton a fine deep-purple shade. boyad-sn} VI. ON THE NITRO DERIVATIVES OF DIPHENYLAMINE. By HUGH RYAN, D.Sc., anp THOMAS GLOVER, M.Sc., University College, Dublin. Read Apri 22. Published Aucust 13, 1918. Tue study of the products formed by the action of nitric acid on diphenylamine, beset as it is by many experimental difficulties, is rendered all the more tedious by the existence in the chemical literature of discrepancies with regard to the polynitro derivatives of the base. Thus 4-nitro-diphenylamine'! is described by O. N. Witt (Ber. d. Dtsch. Chem. Ges. xi (1878), p. 757) as orange leaves melting at 132°C., by Irma Goldberg (D. R. P. 187870) as orange-red leaves melting at 133°C., by E. Bamoverger (Ber.d. Dtsch. Chem. Ges. xxxi (1898), p.580) and F.Ullmann (D.R.P. 193448) as yellow erystals: Picryl-aniline is described as ved prisms by Clemm (Ber. iii (1870), p. 126); and as orange needles by E. Bamberger and Miller (Ber. d. Dtsch. Chem. Ges. xxxiii (1900), p. 108). Similarly for 2.4. 8. 10-tetranitro-diphenylamine Hager (Ber. d. Dtsch. Chem. Ges. xvii (1884), p. 2629) gives the melting-point as 180°C., A. Pictet and E. Klein (Arch. d. Sciences phys. et nat., Genéve (4) xvi (1903), pp. 191-212) as 197° C., Gnehm and Wys (Ber. d. Dtsch. Chem. Ges. x (1877), p. 1819) as 192°C., and P. Juillard (Bull. Soc. Chim., Paris (3) 1905, pp. 1172-1190) as Ge Ce For the removal of the discrepancies with regard to the properties of the nitro-derivatives, and for the establishing of the constitutions of the latter ‘The numerical nomenclature of the diphenylamine derivatives discussed in ‘this communication is based on the formula Ya 3 2 SS a 8 9 SS L4 1S—NU—¢7 10 Ni eA SOIp, ay \ a Ss JES R.1.A. PROC., VOL. XXXIV, SECT. B. [O} 98 Proceedings of the Royal Irish Academy. where doubt existed we have deemed it necessary to prepare, in a pure condition, some of the more important of these nitro derivatives in a manner which would leave no doubt as to their constitutions. We have, in addition, attempted the preparation of some hitherto unknown nitro derivatives of diphenylamine in the hope that a study of their properties may enable us to detect them if present, even in small quantities, amongst the substances formed by the action of nitric acid, at low temperatures and concentrations, on diphenylamine. The various methods which have been recommended for coupling aromatic amines with the halogen derivatives of aromatic nitro bodies have been examined by us, and we have found that in general the best results are got by the method of Irma Goldberg (D.R.P. 185663, 187870; cf. Ber. d. Dtsch. Chem. Ges. x] (1907), pp. 4541-4546), in which a nitrobenzene solution of the components is heated for several hours with anhydrous potassium carbonate aud cuprous iodide, or traces of copper and potassium iodide. In this way we obtained 4- and 3-nitro-diphenylamines which had been previously prepared by I. Goldberg (loc, cit.) and the hitherto unknown 3.10-dinitro-diphenylamine, but attempts to prepare 3.9-dinitro-dipheny- lamine by this method were unsuccessful. 2.4.8.10-Tetranitro-diphenylamine, which frequently occurs amongst the nitration products of diphenylamine, had not been previously obtained by direct conpling of a nitro-amine with the halogen derivative of an aromatic nitro-body. We obtained it by Goldberg’s method from 2.4-dinitro-aniline and 2.4-dinitro-bromo-benzene, a synthesis which removes any doubt that may have existed as to the constitution of the compound. We did not, however, succeed in effecting the combination of picryl chloride and 2.4-dinitro-aniline, by which we hoped to effect a synthtsis of 2.4.6.8.10-penta- nitro-diphenylamine. Although basic amines, such as aniline, combine easily with the nitro- derivatives of bromo-benzene, nitro-derivatives of the amines do not interact so easily with nitro-derivatives of the halogen compound, and in some cases we were unable to isolate any pure product of the reaction. Since, in some cases, it has been observed that a nitro group may be replaced by an amine residue more easily than a halogen atom, we examined the behaviour of B- and y-trinitro-toluenes in reactions such as those we were studying. y-Trinitro-toluene interacts easily in alcoholic solution with m- and p-nitranilines forming well-crystallised derivatives of 5-methyl-diphenylamine, and asmall amount of a crystalline product formed by the action of /3-trinitro- toluene on m-nitraniline is probably derived from the same base. We also examined the action of the oxyacids of nitrogen on some nitro- Ryan And GLtover— Nitro Derivatives of Diphenylamine. 99 derivatives of diphenylamine which are not usually met amongst the products of the action of those acids on the base. 3-Nitro-diphenylamine was converted by nitrous acid into its nitroso derivative, and from the latter by the action of nitric acid in the cold a crystalline compound, which melted at 184°C., and was apparently a trinitro- diphenylnitrosamine, was obtained. On the other hand, 2.4.9-trinitro-diphenylamine was converted by isoamylnitrite in acetic acid solution into a crystalline substance melting at 190° C., the analysis of which agreed more closely with that required for a tetra-nitro-diphenylamine than with that for a trinitro-diphenylnitrosamine. A similar treatment of picryl-aniline yielded two compounds, one of which melted at 236°C., and was probably 2.4.6.8.10.12-hexanitro-diphenylamine, and the other melted at 193-194° C., and seemed to be 2.4.6.8.10-pentanitro- diphenylamine. EXPERIMENTAL. A. Mono-Nitro-Diphenylamines. ie 4-Nitro-diphenylamine O.N .C;5H,. NH. C,H;. We obtained this substance in a manner similar to that by which it had been previously prepared by Goldberg (Joc. cit.). From the parent substances— acetanilide and p-nitro-bromo-benzene—utilised by us in the reaction, the nitro-diphenylamine was formed in good yield. It was purified by erystallisation from alcohol and carbon tetrachloride. From the latter solvent it separated in large, beautiful, iridescent, yellow plates, which became opaque on standing, or more rapidly on being heated. When dry it melted at 153-134° C., and was found to be identical with the product formed by the action of alcoholic potash on 4-nitro-diphenylnitrosamine. According to Goldberg it should give a blue colouration with concentrated sulphuric acid. We found, however, that its solution in concentrated sulphurie acid was colourless, and became intensely violet on addition of sodium nitrite. When warmed with alcoholic potash it formed a blood-red coloured solution. 2! 3-Nitro-Diphenylamine O,NC.H,. NH. C.Hs. We obtained 5-nitro-diphenylamine melting at 112°C. by the method described by Goldberg (/oc. cit.), and converted it by means of nitrous acid into 3-nitro-diphenyluitrosamine. To awell-cooled solution of 4 g. of 3-nitro- diphenylamine and 3 g. of hydrochloric acid (sp. g. 1:19) in a little aleohol a solution of 3:3 g. of sodium nitrite in 6 cc, of water was added slowly with [02] 100 Proceedings of the Royal Irish Academy. constant stirring. The colourless precipitate was filtered, washed free from salt by water, and then washed with a little alcohol and ether. The pure substance, the yield of which was about 80 per cent., consisted of colourless needles melting at 89-90° C., and gave on analysis the following results :— 0:1368 g. substance gave 20-4 cc. moist nitrogen at 22°C, and 764 m.m.p. corresponding to N 17-0 C,:H,O;N; requires N 17°3. 3-Nitro-diphenylnitrosamine consists of colourless, acicular crystals, which are readily soluble in acetone or chloroform, and sparingly in alcohol or acetic acid. Its solution in concentrated sulphuric acid had a violet colour. B. Dinitro-Diphenylamines. 3. 2.4-Dinitro-diphenylamine (O,N),C,H.NH . C,Hs. Following the method of Clemm (Ber. d, Dtsch. Chem. Ges. iii (1870), p. 128) we obtained 2.4-dinitro-diphenylamine by the action of aniline on 2.4-dinitro- chlorobenzene. The product after being washed with dilute hydrochloric acid, was re-crystallised from alcohol and chloroform. It consisted of long orange- red needles, melting at 156-157° C., which were soluble in acetone or glacial acetic acid, slightly in ether or carbon tetrachloride, sparingly in alcohol, and readily in chloroform or benzene. With alcoholic potash it formed an orange-red solution which became blood-red on being warmed, and with concentrated sulphuric acid it gave a nearly colourless solution which became brown on addition of a trace of sodium nitrite. 2.4-Dinitro-diphenylnitrosamine. To a solution of 1 g. of 2.4-dinitro- diphenylamine in 60 ce, of glacial acid, 2 cc. of isoamyl nitrite were added, and the mixture was allowed to remain in a cold room for a few days. The solution, which was at first orange in colour, gradually became much lighter in colour, with simultaneous separation of a pale yellow solid. After addition of water the solid was filtered and washed, first with water, afterwards with alcohol. When re-crystallised from chloroform it melted at 149-151°C., and gave on analysis the following results :— 01029 g. substance gave 16-2 cc. moist nitrogen at 14° C. and 759 m.m.p. corresponding to N 19:2 C,,H.N,Os requires N 19:4. Ryan anp Giover—WMitro Derivatives of Diphenylamine. 101 2.4-Dinitro-diphenylnitrosamine consists of sparingly soluble, pale yellow prisms, which were coloured red by concentrated sulphuric acid, in which they dissolved, forming an orange-yellow solution. Tn another experiment, which was carried out at asomewhat higher mean temperature (ca. 18° C.), the reaction proceeded differently. 2.4.8.10-Tetranitro- diphenylamine separated slowly. When collected and purified, it gave on analysis the following results :— 0:1221 g. substance gave 21°8 cc. moist nitrogen at 20° C, and 760 m.m.p. corresponding to N 20:0 C,.H,O.N; requires N 20:1. 4, 3.10-Dinitro-Diphenylamine O.N .C;5H;,. NH. C,H,. NO. A solution of 6 g. of p-nitro-bromo-benzene and 4 g, of m-nitraniline in nitro-benzene was heated with 2g, of dry potassium carbonate, traces of copper dust and potassium iodide to gentle boiling under a reflux condenser for twenty hours. The nitro-benzene was distilled in a current of steam, and the dark-coloured, tarry residue was extracted with boiling acetic acid. The dinitro-compound, which was precipitated by addition of water to the acetic acid solution, was purified by means of alcohol, and crystallised from chloroform. When dry, it softened at 205°C., melted at 210-212°C., and gave on analysis the following results :— 0°1052 g. substance gave 146 ce. moist nitrogen at 17° C. and 766 m.m.p. corresponding to N 16:2 C,.H,O.N; requires N 16-2. 3°10-Dinitro-diphenylamine consists of pale yellow crystals, which are soluble in chloroform, readily in acetone, and sparingly in cold alcohol. Its solution in alcohol gave a violet red colouration with potash, and that in sulphuric acid, which was colourless, was not affeeted by addition of sodium nitrite. C. Trinitro-Diphenylamines. by, 2.4.6-Trinitro-Diphenylamine (O.N)s3 CsH,NHC,Hs. 2.4.6-Trinitro-diphenylamine, which consists of scarlet red prisms, melting at 178°C., was obtained by the action of picryl chloride on aniline or acetanilide (Clemm, Joc. cit.). 102 Proceedings of the Koyal Irish Academy. - 6. 2.4.9-Trinitro- Diphenylamine (O.N).C;5H;NHC,H,NO2. According to Austen (Ber. d. Dtsch. Chem. Ges. vii (1874), p. 1250) 2.4.9-trinitro-diphenylamine melts at 189°C., and according to Willgerodt (Ber. d. Dtsch. Chem. Ges. ix (1875), p. 1178) it melts at 194-195°C. We obtained it by heating to 170-180°C., for ten hours, an alcoholic solution of 2.4-dinitro-bromo-benzene and m-nitraniline with sodium acetate and cupric oxide. It consists of pale brown, platy crystals, which are difficultly soluble in aleohol, acetic acid, or chloroform, and nearly insoluble in carbon tetrachloride. It melted at 193-194° C. (uncorr.). Its solution in concentrated sulphuric acid had a faint yellow colour which was not affected by addition of a trace of sodium nitrite. 7. 24.9-Trinitro-5-Methyl-Diphenylamine (O,N),(CH;)CsH,.NHC,H,NO,. Was formed by heating 1 g. of y-trinitro-toluene with an equal weight of m-nitraniline in alcoholic solution for two hours. The compound which separated was filtered and re-crystallised from alcohol. It melted with slight decomposition at 199° C., and gave on analysis the following results ;— 0°1672 g. substance gave 26:2 cc. moist nitrogen at 19° C. and 764 m.m.p. corresponding to N 181 C,;H,.N,O, requires N 17°6. 24.9-Trinitro-5-Methyl-Diphenylamine consists of glistening dark-yellow prisms which are sparingly soluble in most organic solvents, excepting acetone, in which they are moderately soluble. Its solution in alcoholic potash has an orange-red colour, while that in sulphuric acid (conc.) is faint yellow, and remains almost unchanged on addition of a trace of sodium nitrite. 8. 2.4.10-Trinitro-5-Methyl-Diphenylamine (O.N)(CH,)C,H.NH . C,H,NO, was prepared by a method similar to that just described, from y-trinitro-toluene and p-nitraniline after heating for thirty hours. After re-crystallisation from a mixture of acetone and alcohol it melted with slight decomposition at 210° C., and gave on analysis the following results :— 01051 g. substance gave 16°6 cc. moist nitrogen at 23°C. and 760 m.m.p. corresponding to N17°9 C,,H,.O.N, requires N 17-6. Ryan anp GLover— Nitro Derivatives of Diphenylamine. 103 9.4.10-Trinitro-5-Methyl-diphenylamine consists of glistening straw- coloured, prismatic needles, which are only slightly soluble in most organic solvents, excepting acetone. The crystals were turned red by contact with concentrated sulphuric acid, in which they dissolved, forming a faintly yellow-coloured solution which remained apparently unchanged on addition of a crystal of sodium nitrite. 9. 2.6.9-Trinitro-5-Methyl- Diphenylanvine (O.N).(CH,)C;H,.NH . C,H,NO:. A substance which is probably 2.6.9-trinitro-3-methyl-diphenylamine was obtained in small quantity by prolonged heating of an alcoholic solution of #-trinitro-toluene and m-nitraniline. It consists of glistening, straw- coloured, prismatic needles, melting with slight decomposition at 199° C., and sparingly soluble in most organic solvents. Its amount was insufficient for analysis. D. Tetranitro-Diphenylamines. 10. 2.4.8.10- Tetranitro-Diphenylamine (O.N}.C,H;. NH .CsH,(NO.).. The symmetrical tetranitro-diphenylamine was obtained in a poor yield: by heating 3 g. of 2.4-dinitro-bromobenzene and 2 g. of 2.4-dinitro-aniline ! in nitro-benzene solution with 2 ¢. of anhydrous potassium carbonate, traces of copper dust, and potassium iodide for several hours to a temperature of about 160°C. The residue left after distilling the nitro-benzene ina current of steam solidified on cooling. It was filtered, washed with alcohol, and re-ecrystallised a few times from glacial acetic acid. Prepared in this way it consisted of brownish prisms which melted at 199-200°C. A mixture of it with an equal amount of the symmetrical tetranitro-diphenylamine obtained by the direct nitration of diphenyl-nitrosamine also melted at the same temperature. 11. = 2.4.6.9- Tetranitro-Diphenylamine (O.N)sCyll,. NH. C;HyNO3. Picryl-m-nitraniline was obtained by Austen (/oc. cit.) by boiling an alcoholic solution of pieryl chloride with m-nitraniline. We obtained it by a similar method, and found that it consisted of short yellow prisms melting at 210° C. (corr.)—a temperature shghtly higher than that (205°) given by Austen, It is not identical with the tetranitro-diphenylamine (MP 190°C.) obtained by the prolonged action of isoamyl nitrite and air (see below) on 2.4.9.-trinitro-diphenylamine. 1 Dinitro-aniline can be conveniently prepared, in almost quantitative yield, by allowing a solution of dinitro-bromo-benzene in alcoholic ammonia to remain in a stoppered flask for a few days at the ordinary temperature. 104 Proceedings of the Royal Irish Academy. ie), 2.4.6.10-Tetranitro-Diphenylamine (O.N);C;xH2NH . C;H\NO:. Picryl-p-nitraniline has been previously obtained by Austen (Joc. cit.) and by Wedekind (Ber. xxxiii (1900), p. 432) by boiling picryl chloride with an alcoholic solution of p-nitraniline. The specimen which we obtained by this method consisted of golden- yellow lozenge-shaped prisms, melting at 222° C., which were slightly soluble in alcohol, ether, benzene, chloroform, or carbon tetrachloride, somewhat more soluble in acetic acid, and soluble in acetone. It gave a red colouration with cold 5 per cent. aqueous alkali, and formed a yellow-coloured solution in concentrated sulphuric acid. E. Action of Nitric Acid on 3-Nitro-Diphenyl-Nitrosamine. To a solution of 1:2 g, (1 mol.) of 3-nitro-diphenyl-nitrosamine in 50 ce. of glacial acetic acid 0°8 ce. (slightly less than 3 mols.) of nitric acid (sp. g. 1-42) was added, and the mixture was allowed to remain 5 weeks at the room temperature. As no solid separated during the reaction, water was added, and the solid, which now separated, was filtered, washed, and dried. On repeated crystallisation from acetone and chloroform yellow prismatic needles were obtained, which softened about 179°C., and melted with decomposition at 184-185°C. An analysis of the substance indicated that it was a trinitro-diphenyl-nitrosamine :— 0:1021 g. substance gave 19-6 ce. moist nitrogen at 24° C, and 746 m.m.p. corresponding to WN 21:2 C,:H,N,0,; requires N 21:0. The trinitro-diphenylnitrosamine consists of yellow prismatic needles which were sparingly soluble in alcohol, were readily soluble in acetic acid or chloroform, and very soluble in acetone. It gave a blood-red colouration with alcoholic potash. F. Action of Nitrous Acid on 2.4.9-Trinitro-Diphenylamine. With a view to identifying of the trinitro-diphenylnitrosamine, described above, an attempt was made to obtain the nitrosamine of 2.4.9-trinitro- diphenylamine by the action of isoamyl nitrite on an acetic acid solution of the latter. The reaction, however, proceeded very slowly, and the product finally isolated in very small quantity was apparently a ¢etranitro- diphenylamine. 0°6 g. of 2.4.9-trinitro-diphenylamine was dissolved in 120 ce. of glacial acetic acid, and to the solution 2 ce. of isoamy] nitrite was added. The solution, Ryan AnD Giovier-—Mitro- Derivatives of Diphenylamine. 105 which at first was amber-coloured, gradually became paler, and a small quantity of a brownish solid very slowly separated. After four weeks water was added, and the solid was filtered, and washed with small quantities of water, alcohol, ether, and chloroform. After several purifications from ‘alcohol and acetone it was finally obtained as yellow crystals, which melted at 190° C,, and gave on analysis the following results :-— 0:0701 g, substance gave 12 cc. moist nitrogen at 18° C. and 764 m.m.p. corresponding to N 19-9 C0,.H,N;O, requires N 2071. Owing to the small amount of substance obtained in the reaction, we were unable to carry out the analysis with a larger quantity than that mentioned. The tetranitro-diphenylamine consists of yellow prisms, which are only slightly soluble in alcohol, soluble in chloroform, and readily soluble in acetone. With alcoholic alkali it formed a blood-red coloured solution. G. Action of Nitrous Acid on Picryl-Aniline. To 7 g. of picryl-aniline, and rather less acetic acid than that required to completely dissolve it, 12 cc. of isoamyl nitrite was added, and the mixture was allowed to remain at the room temperature for four weeks. A small quantity of a yellow crystalline solid separated. After re-crys- tallisation of the solid it melted at 236°C., and probably consisted of 2.4,6.8.10-hexanitro-diphenylamine. The substance contained in the acetic acid filtrate was precipitated by addition of water, and, after repeated crystallisation from acetone and chloroform, it consisted of yellow prisms melting at 193-194° C. An analysis which was made showed it to be a pentanitro-diphenylamine, very probably 2.4.6.8.10-petanitro-diphenylamine, for which P. Juillard (/oe. eit.) gave 195° C. as the melting-point :— 01484 ¢. substance gave 28:0 cc. of moist nitrogen at 20°C. and 757 m.m.p. corresponding to N 215 C,,H,N Oi. requires N 21°35. In conclusion we wish to express our thanks to Nobel’s Explosives Company for a grant in aid of this investigation, and to Mr. Rintoul, the Manager of the Research Section of that Company, owing to whose suggestion the work was undertaken. R,I,A. PROC., VOL. XXXIV, SECT. B. [P] [ 106 ] VEE ON CERTAIN ACTINIARIA COLLECTED OFF IRELAND BY THE IRISH FISHERIES DEPARTMENT, DURING THE YEARS 1899-1915. By T. A. STEPHENSON, Demonstrator in Zoology, University College of Wales, Aberystwyth. (Plates XIV-XX.) (COMMUNICATED BY R. SOUTHERN, B.SC.) Read June 10. Published Novemner 22, 1918. INTRODUCTION. Tue collection of Sea Anemones with which the present paper is concerned has been made by the Fisheries Branch of the Department of Agriculture and Technical Instruction, Iveland, between the years 1899 and 1913. Mr. C. L. Walton, M.Sc., of this College, consented at first to identify the contained specimens; but the collection is a very large one, and Mr. Walton was unable to give to it the necessary time, on account of circumstances which subsequently arose. He therefore handed the specimens over to me, and for some time I have given them my attention. The notes which follow do not deal with more than half the collection at most; but as I shall be unable to deal with the remainder for some time, I am publishing the results so far. 1 should like to say that the descriptions which follow are not intended to present an exhaustive study of the species with which they deal—cirecumstances have not permitted that; but I hope they will be found to give what is essential for identification and general purposes. My original intention was merely to identify the Anemones; but several of them proved to be new species, and some of them to be so interest- ing, that it seeined a pity not to publish some account of them. ‘The micro- scopical part of the work is based for the most part rather on a limited number of selected sections than upon large series, as I have found the latter to be unnecessary for the present purpose. In staining sections of Actiniaria I find the Haematoxylin-Eosin method (see 48,’ p. 2) the most generally ' Numbers in brackets refer to papers in the bibliography at the end of the paper. SrePHENSON— Certain Actiniaria collected off Ireland. 107 useful. But another way which is sometimes very valuable for nematocysts is to stain first for a very short time (not more than half a minute) with methylen blue; then wash quickly in water, stain for an equally short time with picro-satirefuchsin, dehydrate, and clear as rapidly as possible. 1 have only actually used this method once, but it then stained thick-walled sting- cells deep blue, spirocysts vivid purple, and the protoplasm pale purple — the spivocysts being much more clearly visible than otherwise.’ With regard to the Plates, I have drawn them as far as possible in pure black and white to avoid expense; and it must consequently be remembered that this gives to many of the illustrated structures a harder and more definite appearance than they actually have in sections; and this applies, of course, in lesser degree to whole specimens. Finally, I wish to offer my most hearty thanks to Prof. G. C. Bourne, F.R.S., Prof. H. J. Fleure, D.Sc., Dr. E. J. Allen, ’.R.S., and Mr. C. L. Walton, M.Sc., for willing help concerning various points, which they have given me during the course of my work. I should say that Mr. Walton had provisionally identified some of the specimens before I made a start with them, and I am pleased to acknowledge the help which this gave me. DESCRIPTION OF SPECIES. Sub-order ACTINIARIA. J, ACTINIINA M. Edw. Family ILYANTHIDAE Gosse. Actiniina with the aboral extremity of the body rounded, so that there is no pedal disc. No acontia. I am using the term Ilyanthidae here, as before, as a general heading, with certain sub-divisions, but hope to enter into the relationships of the contained forms in another place. Sub-family HALCAMPOMORPHINAE Carlgren. Ilyanthidae with the sphincter endodermal or absent. Group A. PEACHEAE. Halecampomorphinae with 10 pairs of mesenteries only, of which either only 6 pairs or all the 10 pairs may be perfect. Conchula present or absent. Peachia Gosse, 1855. Siphonactinia, Dan. and Kor. 1856, Peachiae with only 6 pairs of perfect mesenteries, and with a conchula and 20 tentacles, which are not knobbed. [P 2) ms . Proceedings of the foyal Irish Academy. 1. P. hastata Gosse. Two specimens. Clxxxii. Bofin. 11. viii. 1900. Port Island Bay; shore collection. There is no need to describe these specimens: the number of tentacles and arrangement of mesenteries refer them with certainty to the genus Peachia. It is a little more difficult to speak certainly as to species, because the conchula is not very well preserved, and it is in any case a structure which may lose some of its characteristics after the animal’s death. But although it has three more or less distinct primary lobes, these are again subdivided, though not deeply; and it approaches much more nearly the conchula of P. hastata than that of undata or triphylla. 1 may add that I have kept living specimens of P. hastata, and that the conchula of that species might well shrink in spirit so that it presented the form shown in the speci- mens under discussion. Eloactis Andres. Peacheae in which all the 10 pairs of mesenteries are well developed and perfect. There is no conchula, and there are no distinct adhesive suckers on the body. Tentacles, 20, with knobbed apices. 2. E. mazeli Jourdan. One specimen. S.R.1171. 19. v. 1911. Eel-trawl. 287-354 fathoms. 46 miles W. } N. of Tearaght Light. Lat. N. 51° 54° 30”: Long. W. 11° 51’ 30”. . _ Miss O. M. Rees has already described the anatomy of this species (45), so I need add nothing. Family ENDOCOELACTIDAE Carlzren. Actiniina with thin or thick, sometimes cartilaginous body-wall, without sphincter or fosse, anid usually with spirocysts in the ectoderm of the body- wall and actinopharynx. Arrangement of mesenteries quite different from the normal Actinian type, owing to the development of the second and third orders of mesenteries in the endocoels. In consequence of this the arrange- ment of tentacles is very different from the normal type. (Among others, 10 tentacles of the first cycle border immediately on those of the second.) Sex organs present on all the stronger mesenteries, including the directive mesenteries. True basilar muscles absent, base present. (Carlgren, modified. See 48, p. 14.) SrepHEenson— Certain Actiniaria collected off Ireland. 109 Carlgrenia n. gen. Endocoelactidae with 6 pairs of macromesenteries bearing circumscribed retractor muscles, filaments, and most or all of them gonads. There are also representatives of 3 orders of micromesenteries ; of these the first order consists of £ pairs, developed in the 4 lateral endocoels, and these 4 pairs have the character of directives, i.e. their longitudinal muscles face away from each other; these 4 pairs run down the whole length of the body, and just reach the actinopharynx at its upper end, but they bear neither developed gonads, filaments, nor retractor muscles. There are other small micro- mesenteries entirely confined to the uppermost part of the body. I have founded the above genus for a small but extremely interesting form which is present in the collection. In a paper (8) on Porponia and related genera, Carlgren made a suggestion which would account for the descent of the 10-rayed forms contained in the Endocoelactidae from a 6-rayed form, by the appearance of +4 pairs of secondary mesenteries with the character of directives, one in each of the endocoels of the 4 lateral pairs of primary mesenteries, instead of, as usual, in a// the exocoels. The abovye- mentioned form presents so interesting a confirmation of this idea, that I propose to honour it by conferring upon it the name of the distinguished zoologist whose work in connection with Actiniaria is so well known. 5. C. desiderata, n. sp. (CAE OGING ines 15 AU DON hires 08 IAL OIG sree le IAL OSG eet 2, G8) Four specimens. SR. 504. 12.ix.07. Trawl. 627-728 fathoms: from Lophohelia. I have been unable at present to work out the detailed anatomy of this form as closely as I could wish, but it is so distinct a species that I think the details which follow will be found to be sufficient for identification ; I hope to publish fuller details at a later date. The specimens were taken from Lophohelia. They are small (the largest measuring, in total length, about 2 cm.), and do not present any striking external features. (Pl. XIV, fig. 1.) Pedal disc present, small, not much exceeding the column. Body not far from cylindrical, inclined to be rather wider above than below, soft in texture. There is a slight marginal parapet and fosse, and sometimes the former is a little notched. The whole surface is wrinkled and somewhat corrugated, but as far as I can see there are no actual verrucae. The whole animal is pale- brown. in spirit, with darker shades on throat, disc, and tentacle bases. 110 Proceedings of the Royal Irish Academy. Tentacles fairly short, stiff, moderately slender; one medium-sized specimen had 38 of them, of rather irregular sizes, the inner on the whole larger than the others. Another specimen had 41. I have not yet worked out exact details as to relationship between tentacles and mesenteries, but judging from the development of mesenteries in the specimen sectionized, the number of tentacles would reach about 68 in the fully grown animal. The throat is rather wide, and the oral disc consequently narrow. There is no trace of a sphincter. ‘The longitudinal musculature of the tentacles and radial musculature of the disc are entirely ectodermal (Pl. XX, figs. 2 and 8), and not very strongly developed. ‘The mesenterial filaments have ciliated lobes. ‘he upper part, at any rate, of the bodywall ectoderm presents an interesting and characteristic feature. It possesses localized batteries of thick-walled nematocysts ; these batteries vary in size and density, but often form an almost solid mass of sting-cells. Between the various batteries the ectoderm typically has these sting-cells sparse or almost absent. (Pl. XVI, fig. 27.) ‘The bodywall ectoderm also contains a moderate, though not large, number of spirocysts, and the bases of the supporting cells are distinctly thickened. ‘There are 6 pairs of macromesenteries, arranged on the usual hexactinian plan, i.e., there are 2 pairs of directives, and 4 ordinary lateral pairs. These macromesenteries all bear very strong circumscribed retractor muscles and mesenterial filaments; and in the larger of two sectionized specimens 10 of the 12 bore well-developed testes; these could not be seen on the other 2, which were damaged, but they may have been present. One specimen showed a curious and presumably accidental feature—one of the macromesenteries had the retractor almost entirely suppressed on the part of the mesentery just above the level of the enterostome; but it was well developed both above and below that level. Beside the circumscribed retractor, the macro- mesenteries (which are, of course, all perfect) have rather feeble parietal muscles, which die off gradually into a fringe of processes between them and the retractor (Pl. XIX, fig. 1). 5 In each of the four lateral endocoels is found a pair of micromesenteries (ie. mesenteries with no developed filament, retractor, or gonad). These run down the whole length of the body, from oral disc to base, but are quite narrow most of the way, bearing simply parietal muscles, whose processes cannot be said to be better developed on one face than on the other, though there seems to be a tendency towards a slightly better development on the side facing the adjacent macromesentery than on the other. but at the very top, just below the oral disc, these mesenteries become just broad enough to reacli the actinopharynx (and so are perfect), and have there a distinct longi- SrrepHenson—Certuin Actiniaria collected off Ireland 111 tudinal musculature on that surface of the mesentery facing the adjacent macromesentery, though it can hardly be spoken of as a retractor. Pl. XX, fig. 9, shows the arrangement of mesenteries on one side of the animal (note that there is 1 actinopharyngeal groove only), and Pl. XTX, fig. 1, shows the muscular portion of a macromesentery and the whole of the adjacent micro- mesentery at a fairly low level. There are, apart from the larger micromesenteries, representatives of two cycles (apparently incomplete cycles) of smaller ones, confined entirely to the underside of the oral disc, and developed of course in the endocoels formed by the macromesenteries with the primary micromesenteries. There is one question which is naturally raised by this species—Is it not simply a young form of some species of Halcurias, and therefore not entitled to generic distinction ? Of course this may be so; but I incline to think that the form is really not very far from adult, and is entitled to generic distinction, for the following reasons. The largest specimen investigated (and not much larger than the others) possessed well-developed testes; and although a small animal, was not much smaller than, for instance, Halewrias pilatus, if at all. Then, if the 4 pairs of micromesenteries are to become macromesenteries—and a great deal of development, above all the specialization of a great retractor muscle, is necessary for the accomplishment of this (a condition which would be required to convert the form into a Halcurias)—one would expect them to be more advanced than they are by the time that the animal has reached its present size and has acquired gonads and tertiary mesenteries. So I have thought it best to erect a new genus for the species, which will stand unless it should be proved in some way that the young Halcurias resembles Carlgrenia. Family BOLOCERIDAE McMuwtrich. Actiniuna with well-developed base and basilar muscles. Sphincter endodermal, variously developed, or absent; if present, may be diffuse or circumscribed. Usually numerous mesenteries perfect. Tentacles attached to the oral dise by a very short thin-walled neck, and usually provided each with a sphincter muscle above the neck, by means of which the neck may be torn, in which case the tentacle is lost. Bolocera Gosse. Boloceridae with well-developed endodermal sphincter-muscle, which is diffuse. Tentacles with well-developed sphincters. Numerous perfect mesenteries (i.e., more than 6 pairs). 112 Proceedings of the Royal Irish Academy. I have narrowed the above definition so that-it excludes B. pollens and B.norwegica. It does not seem to me justifiable to include in one and the same genus species which have a diffuse, and species which have a circumscribed, sphincter; or species which have six, and others which have twelve or more, perfect mesenteries. B. pollens, which has an extremely strong and peculiar circumscribed sphincter, should go to a distinct genus, and might be called Leipsiceras'—a possibility which MeMurrich has already (32) foreshadowed. B. norvegica, Pax, which is a minute but sexually mature species, with only six pairs of perfect mesenteries, can hardly be included in Bolocera. It agrees with Boloceroides in having only six pairs of mesenteries, but whether they have enough else in common, our knowledge of norvegica is not complete enough to show. But it must either go to Boloceroides or to a new genus. The genus Bolocera, as thus limited, will contain :—/3. tuediae, the type; B. kerquelensis ; B. longicornis; B. africana; B. multicornis; B. multipora ; B. pannosa ; B. occidua ; B. brevicornis. I have seen and kept two living specimens which agree perfectly, except for unimportant colour variation, with Gosse’s description of Bolocera eques, which seems to have been considered a doubtful species. Although I have not yet fully investigated these specimens, I may say that the species eques is really an Urticina (‘Tealia), with weak verrucae, circumscribed endodermal sphincter, and non-deciduous tentacles. It is not, I think, identical with Urticina coriacea, the common shore species, but may perhaps be so with Rhodactinia crassicornis. I hope to go further into that question later, but my present point is that it is not a Bolocera. 4. B. tuediae, Gosse. (Pl. XIV, fig. 2; Pl. XX, figs. 1, and 3-6.) (Actinia tuediae, Johnst.; Anthea tuediae, Johust. and others.) SR. 338. 13. v. 06. Trawl. 291-330 fms. Lat. N. 51° 28’ 30”. Long. W. 11° 39’. 1 specimen. SR. 330. 9. v.06. Trawl. 415-374 fms. Lat. N. 51° 16’. Long. W. 11° 37’. One specimen, which has lest nearly all its tentacles. The anatomy of this species has not hitherto been described, and its position therefore has remained a little uncertain. I find in the present collection 2 specimens which I think may be certainly identified as Z. tuediae, from the descriptions of Gosse, Johnston, and Cocks. Mr. C. L. Walton, who has seen living specimens of both this species and of B. longicornis in the North Sea, agrees with me in this respect. 1 Acixw. to release, and «¢eas,a horn. This is founded, like the name Bolocera itself, on the habit of shedding the tentacles. SrepHenson— Certain Actiniaria collected off Ireland. 113 In studying the anatomy of these specimens, however, I find that it presents a close similarity to that of 2. longicornis, and consequently the question arises as to whether the two species are really distinct. Carlgren (1893) has already suggested the possibility of their identity. And, more- over, there are two specimens in the present collection which I cannot, in their preserved’ state, and from external characters, definitely refer to either one species or the other. Nevertheless, I am inclined to think that, in spite of these facts, the two species are distinct, for the following reasons:—When preserved, closely related species of Actiniaria lose their external distinctive characters to a large extent, when there would be no difficulty at all in distinguishing them if they were alive—for instance, from examination of hundreds of specimens in their natural haunts, of Sagartia nivea and S. venusta, I have not the least doubt that they are distinct species; but if they were preserved, it would be difficult, if possible, to distinguish them. Mr. C. L. Walton, who has seen B. tuediae and B. longicornis alive, assures me that in that condition they are quite distinct. With regard to anatomy, one would not expect striking differences in two species which had, possibly, not long diverged from a common origin; and my specimens of &. tvediae show just enough difference to support the idea that it is not identical with, though near, longicornis, 1 do not wish to speak positively on the subject until I have seen the two species alive and healthy, but for the present I prefer to keep them separate. I will therefore describe the external appearance of my specimens, and note how they differ in structure from &. longicornis. They differ in one respect from Gosse’s description. He says there are only 8 cycles of tentacles; but that is not very lmportant, and there are different ways of reckoning the cycles. So, as he does not give the number in each cycle, we may pass this over, although my specimens have more than 5 cycles. External Characters.—(i) The specimen figured in Pl. XLV, fig. 2 :— Greatest diameter of oral dise and tentacles, 9 cm. Length of one of the longest tentacles, 3°9 em. Height of column, 45 cm. Diameter of mouth, 3°8 em.; of column at middle, 53 em.; of pedal dise, 3°53 em. Pedal dise well developed, but of considerably less diameter than the oral disc. Substance of pedal dise and column similar, rather soft; the regular radial striations of the pedal dise go straight on over its margin, and run up the column. Colwnn widening from below upwards, fairly thin-walled, and not stiff. It has longitudinal striations, and transverse wrinkles and furrows, due no doubt to contraction, so that its surface has a somewhat papillose appearance in parts, though there are no real warts or suckers. The margin forms a very slight rim, but it is of the tentaculate type. The prominent R.I.A, PROC., VOL, XXXIV, SECT, B. [Q] 114 Proceedings of the Royal Irish Academy. and apparently non-retractile tentacles leave a fairly wide clear space round the mouth, with its finely ridged lip; the inner ones are large, and are attached to the disc by a narrow neck, above which they suddenly swell out bulbously, then gradually taper off to a blunt point. Round the “neck,” and below the swollen part, is distinctly visible a circular suture, and this is the “line of breakage” of the tentacle. Some of the tentacles aré neatly broken off at this point, and in these the tentacular sphincter (which lies Just above the neck, and is thrown off with the tentacle) is clearly visible as a thin circular veil, projecting across the lumen of the tentacle, and almost block- ing it. The tentacles are stiff, thick-walled (in their present state of con- traction), self-supporting, and distinctly longitudinally fluted. They are a good deal smaller in proportion to the size of the body than in typical specimens of /}. /ongicornis, at any rate. Outer tentacles much smaller than inner. Radii of oral disc not well marked, save among the tentacles. Two actinopharyngeal grooves. Whole animal flesh-coloured in spirit, actinopharynx darker. Tentacles in 5 cycles—12 + 12 + 24 + 48 + ca. 60 = ca. 156. (ii) The other specimen is similar, but has lost nearly all its tentacles, and has the body more puffed out and the wall, in consequence, thinner and more flaccid. The oral dise, moreover, has the same diameter as the pedal. These are only differences which a living specimen would be able to alter constantly. Tentacle-scars in about 4 cycles. Structure.—Specimen (i) has 77 pairs of mesenteries, of which 45 pairs are perfect. The development is somewhat irregular—the first 3 cycles are perfect in varying degrees, the primary mesenterics being joined to the actinopharynx throughout its length, the secondaries running down it less far, the tertiaries still less; the rest do not reach it at all, and this 4th imperfect cycle is not yet fully formed. In one sector of the animal specially examined, the primaries appear sterile, the secondaries have small gonads, and the other mesenteries are very fertile. Specimen (ii) has also 77 pairs of mesenteries in 4 cycles (13p. + 13p. + 25p. +26p.=77.) The first 3 cycles are perfect to a varying extent, and here the 4th-cycle mesenteries also just reach the actinopharynx. (I have shown previously that in &. Jongicornis also, the last cycle may or may not be perfect.) Primaries sterile, at any rate some of them, the others fertile. The mesenterial development is irregular, some of the 3rd-cycle mesenteries on one side of the body being no larger than 4th-cycle ones on the other, and the 4th cycle being incompletely formed. Both internal and external stomata are present. (iii) It is interesting to note that in a specimen of this species from the SireuEenson— Certain Actiniaria collected off Ireland. 11d North Sea, described by C. L. Walton (56), there were also 77 pairs of mesen- teries, and about 127 tentacles, some having been thrown off. It is curious that three specimens of this species should all have 77 pairs of mesenteries only, and suggests that possibly that is about the maximum to be attained in this case. (iv) Through the kindness of Dr. E. J. Allen I have been able to examine externally another North Sea specimen identified as B. tuediae, as regards external characters. This agreed exactly in all essentials with the specimen I have described as (i) above, but had even smaller (proportionately) and blunter tentacles; they were in four cycles only —12, 12, 24, 48; but the whole specimen was smaller than (i). Musculature. —For convenience of reference I will number the three speci- mens with which I have to deal according to the order in which they are above mentioned. (i) is the specimen in Pl. XTV, fig. 2. (ii) is the other Irish specimen, without tentacles. (iv) is the North Sea specimen borrowed from Plymouth. A. Sphincter.—I have figured four sphincters in Pl. XX. Figs. 1, 5, and 6 are from different specimens of B. tuediae (on slightly different scales); fig. 7 is from an Irish B. longicornis. Vig. 1 is from specimen (i), fig. 5 is from specimen (ii), and fig. 6 is from (iv). The sphincter is, of course, endodermal, fairly strong, diffuse, with high, fine, much-branched processes, even tending sometimes to anastomosis. It is difficult to be definite in speaking of the differences between tuediae and Jongicornis, and of course the sphincters of the different specimens of B. twediae vary considerably among themselves, but one may say that (a) all the ¢wediae-sphincters are of one type ; (0) the ¢uediae sphincter has a tendency to be shorter and proportionately broader and rather more definite than in ZL. longicornis (cf. fig. 7), even though this difference is less marked in the “Terra Nova” Jongicornis than in our northern form. The ¢uediae sphincter has a tendency to produce “humps” of mesogloea at different points in its course, but these may be present or absent, and may vary in position from the top end of the sphincter (see fig. 5) to the bottom end (fig. 6), even in the same specimen at different parts of the margin. And the presence of these “humps” does not seem directly dependent on the inser- tions of outer tentacles. The type of sphincter, therefore, seems distinct from the type figured by Kwietniewski for B. kerguelensis, and by MeMurrich for B. pannosa (see 28 and 80), for instance. In these latter the mesogloea at the upper end of the sphineter sends out a kind of twig of mesogloea, which is larger and more branched than the others. In 4. ¢uedive, the “humps” at various levels are hardly of this nature, and may perhaps be due to mere contraction of the wall. (@ 2) 116 Proceedings of the Royal Irish Academy. B. Mesenteries.—On comparing sections of B. tuediae with the sections of my “Terra Nova” Jongicornis, I find that the mesenteries might be described in almost the same terms for both; but the processes of the muscle-pennon are rather more numerous, more crowded, finer, and some of them rather differently branched, in ¢wediae than in Jongicornis; and the parietobasilar muscle of the former is rather the stronger. The nature of a typical tuediae mesentery is shown in Pl. XX, fig. 4. In different mesenteries the parieto- basilar muscle varies, usually forming a distinct free fold. Sometimes one partner in a mesenterial pair is considerably larger than the other. C. Tentacles.—The tentacular longitudinal musculature of twediae (Pl. XX, fig. 3) does not seem essentially different from that of longicornis. One question remains: Is the European B. tuediae identical or not with the American Bolocera described by Verrill as B. tuediae? I cannot give a definite answer. But Prof. McMurrich has favoured me with a sketch of the sphincter of the American form, and though on the whole quite like the European, it has more of an upper “twig” than my specimens show. Sections of a sector of this specimen included representatives of at least 4, possibly 5, cycles of mesenteries, gonads only on those of the last cycle but one. Further research with more material is really required before we can state to what extent the sphincter, for instance, may vary in one and the same species. . Bolocera a ene Carlgren. (Pl. XX, fig. 7.) 1. SR. 172. Nov. 5, 1904. 54 see W. by N. 3 N. Nly. of Tearaght Light. Lat. N. 52° 2’: ‘Ye W. 12° 8’. 454 fathoms. Tow-net on trawl. 1 Tene specimen. 2. SR. 327. May 8, 1906. 60 miles W. } N. of Tearaght Light. Lat. N. 51° 46". Long. W. 12° 14’ 30”. 550-800 fathoms. Trawl. 1 specimen. 3. SR. 506. Sept. 12, 1907. Lat. N. 50° 34’: Long. W. 11°19’. 661- 672 fathoms. Trawl. 1 specimen. 4. SR. 1242. Aug. 14,1911. 56 miles W. 4S. of Great Skellig. Lat. N. d1° 27’; Long. W.11° 55% 1 specimen, with 1 Actinostola. 550-590 fathoms. Trawl. 5. Pigs 1695. Aug. 20, 1913. 514 miles W. 4 N. of Great Skellig. Lat. N. 51° 30°; Long. W.11° 41’. 479-500 fathoms. Trawl. 2 specimens. ad next 3 records must be put down as B. longicornis (7). 6. SR.171. Nov. , 1904. a8 miles N. W. by W. } W. of Tearaght Light. Lat. N. 52° 7’; Long. W. 11° 58’. 337 fathoms. Trawh. 22 specimens, all very battered, with practically no tentacles left, and nearly all turned inside out. SrepHENson— Certain Actiniaria collected off Ireland. 117 7. SR.188. Feb. 3, 1905. 50 miles W. ¢ N. of Tearaght Light. Lat. N. 51° 53’; Long. W. 11°59’. 320-372 fathoms. Trawl. 2 specimens without tentacles, with several Artinauge richardi, 8. SR. 502. Sept. 11,1907. Lat. N. 50° 46’; Long. W. 11° 21’. Trawl. 447-515 fathoms. 5 specimens, with most of the tentacles gone. I having nothing further to add to the history of this species. Family CRIBRINIDAE MeMurrich. “ Actiniina with an adherent base. The column usually more or less verrucose and frequently with acrorhagi at the margin; these, however, never being ramose or frondose. Sphincter endodermal, circumscribed. Usually more than the first cycle of mesenteries perfect. No acontia.” (MeMurrich, 32.) URTICINA Ehrenberg. Cribrinidae without acrorhagi, but with definite verrucae which are not arranged in definite vertical rows as in Cribrina, but are irregularly scattered over the surface. Numerous (more than six) pairs of perfect mesenteries. Longitudinal musculature of tentacles mesogloeal. 6. U. coriacea Ciivier. (For synonymy of this species and genus see Carlgren, 1893, p. 58.) 1. W. 131. March 11, 1910. Blacksod Bay, Feorinyeeo Bay, North. Shore collection. 2 specimens. Anthopleura Duch. and Mich. Cribrinidae with definite acrorhagi containing batteries of nematocysts. Verrucae present, and arranged in definite vertical rows. Numerous perfect Y to} mesenteries. Longitudinal musculature of the tentacles ectodermal. 7. A. alfordi Gosse. (Aegeon alfordi, Gosse, 1865.) 1. St. vii. February 28, 1899. Ballynakill, vii. Coastguard Bay. Shore collection. 8 specimens. 2. March 3, 1899. On the Bar, Fahy Bay. 1 specimen. This species has already been described by Miss O. M. Rees (47), and | have very little to add. One of my specimens has produced a small lateral bud which has developed tentacles. I can state that the large tooth-like marginal verrucae are truly acrorhagi in that they contain a battery of nematocysts. There is one point in which I should like to add to Miss Rees’ description. My sections show the character of the sphincter to be, typically, 118 Proceedings of the Royal Irish Academy. rather different from the case shown in her fig. 11, which was made under difficulties, as the specimen was not weli preserved, and is, I think, exceptional. When more typically shown it is not a strong sphincter, if we compare it, for instance, with the powerful muscle of Epiactis novo-zealandica (48, Pl. V1, fig. 3), and it is more attached to the wall than Miss Rees’ figure indicates— not merely by a narrow pedicel, but by a broader area of attachment. Family PARACTIDAE' R. Hertwig. Actiniina with well-developed base and basilar muscles. Sphincter mesogloeal. Acontia absent. Sub-family ACTINOSTOLINAE! Carlgren. Paractidae in which, in the younger cycles of mesenteries, the individual mesenteries of each pair are unequally developed, exactly according to the following rule:—The larger mesentery of each pair (ie. the older) has its longitudinal retractor muscle turned towards the nearest mesentery-pair of the next oldest cycle. Actinostola Verrill. Actinostolinae with smooth or tuberculated bodywall, which may be very thick. The oral dise may be more or less undulated in outline. Margin tentaculate. Tentacles numerous, in more than two cycles, with or without basal swellings of mesogloea on the aboral side. Longitudinal musculature of tentacles and radial musculature of oral disc entirely embedded in the mesogloea. Numerous perfect mesenteries, the primaries at least, and usually more, sterile. Pennons diffuse. 8. A. atrostoma n. sp. (PL XIV, figs 5, 7,8; PL XV, fig. 7; PL XVI, figs. 11, 12, 16-20; Pl. XVII, figs. 1-4, 6-8, and 17.) 1. SR. 487 and 489. 4 large specimens. — SR. 487. Sept. 3, 1907. Lat. N. 51° 36’; Long. W. 11° 57’. Trawl. 540-660 fathoms. SR. 489. Sept. 4,1907. Lat. N. 51° 35°; Long. W. 11° 5’. Trawl. 720 fathoms. ' I should like to state here that in this paper I am only using the families Paractidae and Sagartiidae, the sub-families Actinostolinae and Paractinae, and the genera Cymbactis and Actinauge (as those families and genera are at present understood) proriswnally. 1 have been working for some time at a scheme for a somewhat different grouping of all the genera at present contained in the families Paractidae and Sagartiidae, and hope to pub- lish it shortly. But as it is not convenient to enter here into the long discussions involved, I am retaining the present classification in the meanume. Srepnenson—Certain Actiniaria collected off Ireland. 119 2. SR. 1242. Aug. 14,1911. 56 miles W.}5S. of Great Skellig. Lat. N. 51° 27’; Long. W. 11° 55’. 550-590 fathoms. Trawl. 1 young specimen with a Bolocera. 3, One large specimen, with no label. Measurements.—(i) Largest specimen. Diameter of oral dise and tentacles, 12em.; of mouth, 5°3 em.; of pedal disc, 45 cm. ; height of column, 7°5 cm. ; length of .a large tentacle, 2 cm, (ii) Smallest specimen. Diameter of oral disc and tentacles, 6 cm. External characters.—Pedal dise in all cases much smaller than oral dise, so that the body is cup-shaped. Pedal disc flat or hollow, round or irregular, much cockled, more or less radially striate. Form of body variable according to state of expansion or distortion. (See Pl. XIV, fig. 5.) Limbus well marked; margin tentaculate, the outer tentacles, which are mere knobs,- passing into it. Bodywall very thick and hard, thickly covered in its upper part, in all cases, with irregular tubercles and prominences. In some cases these extend almost to the limbus, getting smaller and less marked as they approach it; in others the lower part of the body is almost smooth. In the largest specimen the oral disc is widely expanded, and is purely oval in shape. In the other cases, where it is more or less folded up, it forms a deep cup, and its outline is rather irregular or wavy, though not definitely lobed. Dise very broad, thin-walled. adii distinct, corresponding to the larger tentacles, not reaching the mouth, and clearly transversely striated. Mouth large, with prominent lip, and always two well-marked actinopharyngeal grooves. The tentacles (Pl. XIV, figs. 7,8; Pl. XV, fig. 7) are arranged in 5 cyeles—on the plan 6, 6, 12, 24, 48, 96 = 192, this formula sometimes being departed from in small details. In the largest specimen part of a 6th cycle has appeared. The twelve innermost tentacles stand out clear on the dise, and the others are arranged between them in 12 little wedges, the apices of the wedges pointing toward the mouth. Inner tentacles much larger than the outer, and here and there one or more tentacles are aborted, making the size rather irregular. Each tentacle is provided with a hard whitish swelling of the mesogloea on the outer side at the base, very well marked where best developed. The tentacles themselves are thick-walled, provided with distinct terminal pores, and are longitudinally and transversely wrinkled. The body is creamy white except in the smallest specimen, in which it has a pinkish tinge. In four specimens; the dise andsoft parts of the tentacles are a vich deep purple-brown ; in another the colour dies off to paler greyish- brown on the tentacles. The largest specimen is a lighter grey-brown, but looks as if it may have originally been darker. The internal parts of the body are in all a deep rich brown, except for the pale yellow gonads. The 120 Proceedings of the Royal Irish Academy. animal as a whole presents a superficial resemblance to Hertwig’s Polysi- phonia tuberosa, but the tentacular plan, as well as other details, is quite different. Structure.—(i) Mesenteries. 1 carefully dissected three specimens. There are five cycles of mesenteries. and faint traces of the beginning of a sixth in the largest specimen. The first two cycles are quite perfect, the third-cycle mesenteries run less far down the actinopharynx, and the fourth cycle is only just perfect. Total, 96 pairs; arrangement, hexamerous. Tirst two cycles sterile, third with a few gonads, fourth and fifth very fertile. Small oral stomata are present, but no marginal ones. In two specimens, the fourth-cycle mesentery-pairs all had one individual quite clearly larger than the other, and in those of the fifth cycle the difference in size in each ‘pair was extremely marked. The inequality in all cases, and with no excep- tions, followed the rule laid down for this genus by Carlgren in 1893, and which is given above in the definition of the sub-family. In the third specimen, the smaller partner of each fifth-cycle pair was too small to be seen with the naked eye; but by microscopic examination I satisfied myself that, at any rate in the piece of the animal examined, the smaller partner was really there in the right position, but had not yet grown out beyond the endoderm, so no doubt the rule prevailed here as in the other two specimens. The inequality of the mesenteries is not confined to the fourth and fifth cycles, but is less marked in the others. Sections of a sector of the animal reveal the following details :— Fifth Cycle—The smaller partner in each pair has or has not a filament, but it has a small parietal muscle, although no gonad or definite retractor. The larger partner bas a large gonad and filament, with a fringe of stout longitudinal muscle-processes on the endocoelic face, and a streak of trans- verse muscle on the other. The fourth-cycle mesenteries bear large gonads, but the muscular portion of one partner much exceeds that of the other. They have a parietal muscle, tending to be differentiated into parietobasilar and longitudinal halves, and a distinct diffuse pennon, which ceases abruptly at its distal edge. As the body-wall is approached its processes become gradually lower and stout instead of slender, branched, and tree-like. The mesogloea of the mesen- teries, here and in the older cycles, is well developed. The third-cycle mesenteries have the whole of the endocoelic face of their muscular part occupied by a diffuse pennon, whose processes start imme- diately at the body-wall, and rapidly become high. It tapers at both ends, the highest part being nearer the distal side than the proximal. In its best- developed part the mesogloea of the mesentery is produced into rather SrepHunson — Certain Actiniaria collected off Ireland. 121 regular mounds, round which the slender branched processes of the pennon are arranged. (Pl. XVII, fig. 6.) The exocoelic surface is fringed by a straight live of transverse fibres, except for a short distance close to the body-wall where there is a parieto- basilar muscle, with short, rather stout processes. It may end as a distinct fold or not. It seems, as it grows, to enclose in the mesogloea of the mesentery little spaces containing remnants of endoderm, and these are fringed with muscle fibres. The larger mesenteries differ from these only in size. (ui) Sphincter.—Well developed, moderately long and strong. (Pl. XVII, fig. 4.) It lies immediately against the endodermal side of the mesogloea throughout its length, and, except at its widest point, only occupies a small proportion of the thickness of that layer. ‘lhe widest point is near its upper extremity, and from that it tapers off very slowly and gradually downwards. The lower portion is a clearly defined band, which passes directly into the endodermal circular muscle of the body-wall on its inner side. his part is clearly and beautifully reticular in structure (Pl. XVII, fig. 17), except towards its outer edge, where it tends to become looser, and so alveolar. The meshes of the network are fine and very clear. In some sections there is a definitely reticular inner band next to the endoderm throughout the length of the sphincter; but in all cases either the outer part alone, or else the whole width of the muscle, splits, as it goes upwards, into slightly radiating, sub- parallel vertical layers, separated more or less from each other by clear mesogloea. These layers are alveolar in structure, and not very definitely bounded (Pl. XVII, fig. 5), but they vary in width and clearness of definition in different sections. (iii) Dise and Tentacles—The radial musculature of the disc is well developed, and entirely embedded in the mesogloea. It varies considerably, however, in sections taken from different parts of the dise. One set of sections shows it as an almost continuous band of varying width (Pl. XVII, fig. 8) near the middle of the mesogloea. The outlines of the band are not very distinct, and sometimes it is broken up into large clumps, sometimes split into two different bands. It is either “alveolar” or “reticulo-alveolar” in structure, the alveoli on the whole being small. The band is slightly interrupted above the insertions of some of the mesenteries, not perceptibly above others. Then in other sections it is much better developed, definitely reticular in structure, with very large meshes (see Pl. XVII, fig. 7, and note that it is on approximately the same scale as fig. 8), and is sharply inter- rupted over the insertions of some of the mesenteries. In still other sections the meshes are intermediate in size Unfortunately these three sets of R.I.A. PROC., VOL. XXX1V, SEOT. B. [R] 122 Proceedings of the Royal Irish Academy. sections got mixed; but I am inclined to think that the larger meshes represent the outer, the smaller the inner, part of the disc. In the tentacles the longitudinal musculature is also entirely mesogloeal, and runs in a narrow, almost straight band through the mesogloea, fairly near the endoderm. (Pl. XVII, figs. 1, 2.) It is better developed in some places than in others, and sometimes more wavy than others. In structure it is alveolar, sometimes tending to be reticular, and is often only a single chain of alveoli. These latter vary a good deal in size, some being fairly large. At the bases of the tentacles, where the sections pass through the basal swelling (Pl. XVII, fig. 1), it is seen that as the band of muscle reaches the swelling it thins out rapidly, so that at the back of the swelling it is reduced to practically nothing. The mesogloea of the tentacle is thicker than either of the other layers, and its outer side is much folded, in con- sequence of the surface-wrinkles,.and of course the ectoderm follows its out- line. The basal swelling has a coating of ectoderm; but it gets thinner and thinner as the back of the swelling is reached, and there it is very low and poor. (iv) General histology.—The specimen sectionized is a male. Well- developed ciliated lobes are present in the mesenterial filaments, and have their nuclei mainly concentrated in the inner parts; the glandular lobes contain many coarsely granular gland-cells. The whole endoderm in this species is rather particularly interesting, and contains elements of unusual appearance whose detailed study must be left for the present. Among otlier things, there are objects which appear to be large and curious parasitic algal cells with a shining test and possibly a spiral thickening in the wall. The tentacle-ectoderm is well developed, with scattered pigment and very numerous spirocysts, which, though not very broad, are long—often nearly as long as the depth of the ectoderm. Nerve-layer distinct. Gland-cells present in the inner part; also a few of the presumed parasitic algae. The ectodermal margin of the mesogloea appears to have a feeble fringe of muscle fibres, quite distinct and apart from the real longitudinal musculature ; but I cannot be certain about this point. The ectoderm contains narrow, thick-walled cnidae, longer than those found in the body-wall. The ectoderm of the body-wall is low, not pigmented, with small scattered nuclei. A good many quite short, rather narrow, thick-walled cnidae (Pl. XVI, fig. 11) are present, aud one sees even, very rarely, a spirocyst—though these may have stuck to the ectoderm surface accidentally, coming from some other part. There are numerous large, transparent, apparently glandular cells with fine granules which absorb eosin. (v) The only other species of the genus Actinostola beside this one, which have basal swellings to the tentacles, are A. Carlgreni, Wass., and A. intermedia, Srepuunson—Certuin Actiniaria collected off Ireland. 123 Carlgr. A. Carlgreni differs in various small ways from A, atiostoma—it has a different sphincter, stronger tentacle-musculature, and other differences— see Wassilieff 1908. Carlgren gives no figures of A. intermedia (10), but judging from his description, and from comparison with his figures of other species to which he refers, it would seem that the sphincter, disc-radial- musculature, ete., present numerous slight but sufficient differences ; more- over, in A. intermedia, the longitudinal musculature of the tentacles is better developed in the outer basal parts than in the inner. ‘lhis is exactly the opposite in A. atrostoma. Sub-family PARACTINAE! Carlgren. Paractidae in which the two mesenteries of one and the same pair, in the younger cycles, are either equally developed, or, if any inequality occurs it shows considerable irregularity, or does not follow the Actimostola-rule exactly. Cymbactis! MeMurrich. Paractinae with thick smooth body-wall, without verrucae, collar, cuticle, or true capitular ridges (though the upper part of the column may be thrown, in strong contraction, into ridge-like folds), Margin tentaculate, not lobed. Tentacles simple or somewhat thickened at the base, in more than two cycles, their longitudinal musculature mesogloeal. Sphincter simple. More than six pairs of mesenteries perfect. Pennons diffuse. I have been obliged to extend this definition even more than before, to include the rather curious form described below. 9. C. Gossei, n. sp. (PBL XIV, figs: 10) 15 Pl. XVID, figs. 5, 9,14.) SR. 335. May 12, 1906. Lat. N. 51° 12’ 30” — 51° 17’ 30”; Long. W. 12° 18’ - 12° 16’. Trawl. 893-673 fathoms. 1 specimen (with several Actinernus aurelia). Measurements. —Diameter of mouth of basal cavity, 3°6 em.; of column, 65 cm. at greatest; height of specimen (contracted), ca. 5°5 em. Depth of basal cavity, ca. ‘7 em. External characters—The body is almost spherical, having the appearance of an apple with a broad base, the tentacles and upper part of the body being completely introverted. The body-wall is thick throughout, and of hard and firm texture. At the limbus it becomes much thinner, and the edge of the ' See foot-note on p. 118. [R 2] 124 Proceedings of the Royal Irish Academy. limbus is drawn inwards on all sides, so that it forms a perfectly smooth circular rim, the diameter of the circle being much less than that of the body. This encloses a shallow basal cavity (see Pl. XIV, fig. 10), whose diameter is larger inside than at the mouth. The pedal disc itself, which, of course, forms the inner wall of the cavity, is very thin and membranous, and the mesenterial insertions show through it. In the upper part of the body the wall increases in thickness, and at its thickest may be as much as ‘8 cm. The whole upper portion is introverted, and when such a thick, solid substance is squeezed together by such strong contraction, it naturally falls into folds and creases; these might be termed “capitular ridges”; but I am inclined to think, since they are of extremely variable size, and absolutely without any regular arrangement, that they are merely contraction folds and would very likely vanish if the animal were expanded. They are solid mesogloea. The tentacles are in three apparent cycles at the disc margin, and the outer ones pass directly into the folds of the body-wall, so that there is no definite margin. The tentacles are short, stumpy, and of very irregular sizes; the mesogloea of the lower half is very thick and stiff, especially on the outer side, and the upper part is flaccid and much more thin-walled, often with a somewhat capitate lip. (Pl. XIV, fig. 11.) They are peculiarly arranged—the outer ones are much mixed up with the marginal contraction-folds, and the inner ones are just free from the margin. The plan of arrangement, apart from minor irregularities, seems to be 36 + 36 + 72 = 144 (4+ 7 odd ones). The oral disc is smooth and firm, with a slight radial ridge for each tentacle, which does not reach the mouth. Lip prominent. Two well-developed actinopharyngeal grooves. Colour in spirit, white, dise and oesophagus tinged brown. Structure.—(i) Mesenteries. These are arranged on the plan 18p. + 18p. + 36p. + 72p. = 144p., but there are actually 19 pairs in the second cycle, which causes a little irregularity—apart from which the plan is almost exactly carried out. 36 pairs are perfect. The last cycle mesenteries are very small, and alone bear visible gonads. Small oral, but no marginal stomata, are present. A certain amount of inequality in the development of the mesenteries occurs; but altogether, as many pairs have the partners equal (to the naked eye at all events) as unequal, and when one partner is larger than the other it is sometimes the wrong partner that is larger—i.e., it violates the rule which obtains in Actinostola, The fourth-cycle mesenteries bear male gonads, but have no musculature to speak of. The rest of the mesenteries are thin. Those of the third cycle are fringed along the endocoelic side with short, simple, or slightly branched longitudinal-muscle processes, which do not attain the SrepHenson— Certain Actiniaria collected off Ireland. 125 distinction of a pennon. The larger mesenteries (1st and 2nd cycles) have a similar fringe in their proximal part, but in the distal part it rises gradually, till it forms a feeble pennon, with moderately high processes, which are never much branched, and often simple. It ends abruptly, and leaves a narrow non-muscular part behind the filament. (PI. XVII, fig. 14.) Here and there alittle muscle is embedded in the mesogloea. The parieto- basilar muscle, at the level of my sections, is weak, and extends over the greater part of the opposite surface of the mesentery to the pennon. Its distal ending varies, even in two mesenteries of the same pair—it may taper away, or form a slight hump, or even a small free fold provided with a few little stumpy processes. (ii) Sphineter—Fairly large. It must be very strong to be able to contract enough to close up such a stiff animal. It is long (Pl. XIV, fig. 10, shows its position), and lies throughout against the endoderm, It is at its best develop- ment about the middle, being very irregular above on account of the folds into which the wall is thrown, and their continuity with the tentacle-bases. Below it rapidly narrows into a long tail. This lower part is cleanly marked off from the rest of the mesogloea, and is reticular (Pl. XVII, fig. 9) in structure, the meshes being often large, with a tendency to transverse elongation, and fine. From about the middle upwards the reticular structure becomes gradually alveolar, more and more mesogloea intervening between the muscle-cavities. The upper part is poorly and irregularly developed, the scattered cavities varying very much in form. In addition to a tendency to transverse elonga- tion of the cavities, there is a distinct but vaguely marked vertical layering. One can hardly tell where the body-wall ends and the tentacle-bases begin. (iii) Dise and Tentacles—TVhe radial musculature of the disc is much better developed in the sections I have than the longitudinal musculature of the tentacles. It forms a well-defined band of varying width, embedded in the mesogloea, and at its best not occupying more than about $ of the width of the mesogloea. It lies sometimes about the middle, sometimes near the ecto- dermal side of the mesogloea. Where best developed the muscle cavities are close enough together to give the band a reticular structure, but in the thinner places the cavities are more scattered, and the band is only one or two cavities deep. It is usually interrupted or thinned out above the insertions of larger mesenteries. The mesogloea of the tentacles is extremely thick, espeeially at the base, where it is almost equally thick all round in some places. The mesogloeal longitudinal muscle (Pl. XVII, fig. 5) is very feeble, especially near the base, where it lies near the ectodermal surface, often as a single row of cavities. Even in the thin part of the tentacle it seems better developed on the inner than the outer side, 126 Proceedings of the Royal Irish Academy. (iv) General histology—The ectoderm, as far as the portions which I sectionized are concerned, seems to have almost entirely vanished, so that I can say nothing about it. The mesenterial filaments have well-developed ciliated lobes, which, in a simple trefoil, are similar in size and shape to the glandular lobe, and when large are folded. The glandular lobes contain a fair number of narrow, medium-sized, blunt-ended, thick-walled cnidae, and a few coarsely granular gland cells. (v) Cymbactis gossei is distinguished from other species in the genus by the curiously thickened bases to the tentacles, and also by differences in its sphincter and mesenteries and dise-radial-musculature, for the details of which reference should be made to the descriptions of several species given by other authors. (80 and 57.) I have named it after P. H. Gosse, because, as far as I am aware, no Actinian bears the name of the well-known zoologist whose “ Actinologia Britannica” was really the foundation of the serious study of Actiniaria. Stomphia Gosse. Paractinae (?) with body-wall of varying thickness, devoid of verrucae, collar, ridges, &e. Tentacles, disc, and sphincter simple. Numerous perfect mesenteries with diffuse pennons. Longitudinal muscles of tentacles meso- gloeal. More than two cycles of tentacles. 10. S. churchiae Gosse. 1. R10. May 3, 1905. Nymphe bank. 41-42 fathoms. 15 miles off Mine Head. 1 specimen. 2. SR. 1358. May 6,1912. 12 miles S. by W. 4 W. of Chicken Rock Light. Lat. N. 53°50’; Long. W.4°51’. Dredge. 45 fathoms. 4 specimens Carlgren has already fully described a single specimen of this species (6), so I will merely note one or two points with regard to the five further speci- mens in this collection. The mesogloeal sphincter agrees well with Carlgren’s figures (Taf. IX, figs. 2 and 3); the mesenteries also agree essentially, though in my specimen the mesentery contains rather less mesogloea, and the processes of the pennon are proportionately higher—probably a merely individual variation. One of my specimens has the tentacles 16 + 17 + 33; another has them 16 + 16 + 32; a third, 13 + 15 + 31 (here two primary tentacles are absent, if not three, and the whole of that sector of the disc is aborted}; the fourth, 16 + 15 (one absent, but its radius present ; two of the primary 16 aborted) + 32; and the fifth 15+16+31. From this it would seem that an octamerous plan—16, 16, 32—is typical of the species, though not always carried out SrepHiunson—Certain Actiniaria collected off Lrelund. 127 , quite exactly. Carlgren’s specimen had 6 — 12 - 18 - 36; and Gosse gives 6-6 -12 - 36. In two dissected specimens, one had 16 pairs of perfect mesenteries, the other had 1 extra perfect unpaired mesentery. Both had a cycle of moderate-sized imperfect pairs alternating with the perfect ones, and a last cycle, consisting mainly of small unpaired mesenteries, only well developed near the base, and sometimes with a still smaller partner. After careful investigation of the whole of these two animals, I cannot at all satisfy myself that the rule which prevails in connexion with the smaller mesenteries of Actinostola is carried out with any definiteness. It does hold good for some sectors of the animal, but is not, as far as I can tell, general. It seems to me too imperfect, here, to warrant the inclusion of the genus in the Actinostolinae. Actinernus Verrill. Paractinae, with pedal dise of variable extent—it may be broad, or so reduced as to be practically absent; or it may be concave, secreting a cuticle and enclosing mud. Column with a smooth wall, which is sometimes so thick that it forms a jelly-like coating to the body, so that the general form may resemble that of a Scyphozoan. ‘lentacles arranged in two cycles only, at the extreme margin of the oral disc; they all have aboral basal swellings of mesogloea, which are continuous with the body-margin, at least in the case of the outer ones. Longitudinal muscles of the tentacles and radial muscles of the dise entirely ectodermal. Oral dise wide, sometimes lobed. Only six pairs of perfect mesenteries, with weak, diffuse retractors. The species which have hitherto been described under the generic name Actinernus are four in number. The type is A. nobilis, Verrill, and it is unfortunate that its anatomy is unknown. Carlgren (8), from an external examination of a specimen of it, is inclined to think that it may really be a Porponia. A description will be found below of an anemone which I believe to be identical with the second species deseribed—A. sayinatus, Verrill, 1882. This is certainly not a Porponia, since it has the normal mesenterial arrange- ment, only six pairs of mesenteries perfect, and a mesogloeal sphincter. The anatomy of A. plebevus, MeMurrich, 1893, brings it into line with A. saginatus, and I am now able to describe a new species (4. aurelia), with the same essential structural features. The anatomy of A. sp., Haddon, 1888, is unknown. If it should prove in the end that A. nobilis is a Porponia, some readjust- ment of names may be necessary; but in any case, A. saginatus, plebeius, and aurelia go together and form one good genus. MeMurrich (1893) 128 Proceedings of the Royal Irish Academy. ~ suggested that Polysiphonia tuberosa, Hertwig, might be an Actinernus; but it is really quite a distinct form, and more related to Actinostola, as I hope to show in another place. 11. A. saginatus Verrill. (Pl. XV, figs. 4and 9; Pl. XVI, figs. 1-10; Pl. XVIIL, figs. 2-7 and 10.) Helga exx. Aug. 24, 1901. 77 miles W.N.W. of Achill Head. 382 fathoms. 1 specimen. Measurements—Diameter of the basal fissure when opened, 1-7 cm.; of contracted oral disc, 5°5 em.; of expanded disc and tentacles, 11:3 cm.; length of basal fissure, 7-4 cm.; of mouth, 4cm.; of a large tentacle, 2°7 cm. ; height of column, 6 em. External characters.—Vedal dise formed into a long, deep, and narrow crack, which embraces a long spine belonging to some other animal. Con- sequently the pedal disc is as long as the diameter of the body, but not nearly as wide; its edges wrap themselves round the spine, and meet each other, but do not fuse save at one end. The body is whitish, and perfectly shapeless, the oral disc and tentacles being partly incurved and hidden. Its wall forms a soft, thick, jelly-like coat, like the test of an Ascidian, which is smooth but for irregular pits and furrows. The margin is continuous with the basal swellings of the outer tentacles. The long, slender tentacles, tapering gradually to a point, and the disc, are pale salmon. The tentacles are stiff enough to support themselves, and the rather thin dise has, as usual, radii corresponding to the tentacles, which do not reach the wide mouth, with its soft lips. As the animal is preserved, the disc is infolded, but it can be opened out, and it then exceeds the column, and, though not truly circular in outline, is only rather irregularly wavy, not definitely lobed. The tentacles are 137 in number, and are arranged in two alternating cycles only (sixty- eight marginal, the rest non-marginal) at the extreme edge of the disc, even the inner ones communicating more or less with the margin. It seems to me that in this case it would be imagination to divide the inner tentacles into more than one cycle. ‘They are all, allowing for the undulate outline of the disc, approximately on one level, and all sub-marginal. All the tentacles have basal swellings of mesogloea (PI. XV, fig. 9), which vary in — size in individual cases, but are on the whole larger on the outer than the inner ones. The tentacles themselves are rather uneven in size, a few being aborted, but are on the whole sub-equal. (For general appearance, see Pl. XV, fig. 4.) Structure: (i) Mesenteries—Only six pairs of mesenteries are perfect, and even these are not all equally so, There are five cycles (sixty-seven pairs), and STwPHENSON— Certain Actiniaria collected off Ireland. 129 mesenteries of all cycles, save the primary, are fertile; the last (fifth) eyele is not fully formed. There seem to be oral, but no marginal stomata. There is a tendency towards one mesentery of a pair being larger than its partner in the younger cycles, but it is not well or clearly developed, and in the majority of pairs it would not be possible to say which was the larger, by dissection at any rate. Sections of a sector show that the first three cycles bear weak pennons, the fourth and fifth being reduced to little more than reproductive septa. The mesenteries are very thin, and all have a slight thickening of the mesogloea, and a parietal muscle where they join the body- wall. (Pl. XVIIL fig. 10.) (a) A typical directive mesentery.—The parietal muscle has the processes on the longitudinal-muscle side larger, fewer, and more branched than on the other. Most of the mesentery is very feebly muscular, though there are short, stout processes on the endocoelic side. But as the edge of the mesentery which joins the actinopharynx is reached, a weak diffuse pennon appears, its processes becoming higher and higher till they abruptly end at the juncture. They are very stout and little branched, and crowded, so that they have a somewhat “reversed” appearance (¢/. Pl. XVIII, fig. 3) (7e., as if the pro- cesses projected from endoderm to mesogloea, instead of the reverse, which is really the case). As to the rest of the mesenteries, the parietal muscle is always better developed on the same face of the mesentery that has the pennon than on the other. The greater part of the mesentery is thin, and feebly muscular, the second and third cycles bearing a feeble pennon at the distal border (at the level sectionized), which occupies only a small proportion of the muscular surface, and tapers at both ends, typically more abruptly distally. The processes are short and stout, and not much branched, giving almost the “reversed” appearance referred to above. The parieto-basilar muscle, at the level sectionized, is weak, and has no free edge, but fringes almost the whole muscular part of the mesentery, often rising into lobe-like processes. (ii) Sphincter.—Mesogloeal, rather weak. (Pl. XVIII, fig. 6.) It is narrow throughout, widest somewhat below the top, the upper part being mixed up with the tentacle-bases, and so variable in different sections. At its lower end it lies close to, but never in contact with, the endoderm, from which it is separated throughout by a band of mesogloea; above, it approaches the ectodermal surface of the mesogloea. 1t ends rather abruptly, both above and below—below the muscle bundles become fewer, smaller, and more scattered; above (Pl. XVIII, fig. 5) the terminal bundles are distinct. In structure it is alveolar, the fibres to a large extent being arranged in little rings of variable shape and size as to detail. Throughout the greater part of R.I.A. PROG., VOL. XXXIV, SECT. B. [S] 150 Proceedings of the Royal Irish Academy. - the sphincter the alveoli are very small (Pl. XVIII, fig. 2), and are arranged some singly, some in clumps of varying size, loosely or compactly. But at the uppermost end of the sphincter the alveoli are larger and close together, so that the clumps form networks. The sphincter does not exhibit vertical or transverse layering. (iii) Oral disc and tentacles.—The radial musculature of the disc is entirely ectodermal, and is not much different from the tentacle-musculature. The tentacles have thick but variable mesogloea, and high ectoderm, which, how- ever, becomes poorer on the thickened outer side of the tentacle, especially at the base. The entirely ectodermal longitudinal musculature forms a dense fringe round the edge of the mesogloea, varying in width in different parts. The processes are fairly stout and not greatly branched, are moderately high in general, and crowded and rather moss-like in appearance. (Pl. XVIII, figs. 4 and 7.) The musculature almost dies away on the aboral side of the tentacle as the base is approached, and at the back of the basal swelling is hardly present. (iv) General histology—t(a) Ectoderm. In sections of the actinopharynx this is very high, and is crowded with large granular gland-cells; there are also other gland-cells, rather like a mass of small bubbles in appearance, which absorb eosin, and a fair number of moderately long and narrow thick- walled nematocysts, and even a spirocyst here and there. The tentacle . ectoderm is crowded with spirocysts of all sizes, the largest being unusually big. There are also thick-walled blunt-ended cnidae, longer and more cylindrical than those of the mesenterial filament. Nuclei are most plentiful in the middle of the ectoderm. There are a large number of gland-cells of a peculiar sort which I have not seen elsewhere. (Pl. XVI, figs. 5,30.) The body of the cell is stained pale-pink with eosin, and it contains black haematoxylin-stained granules which are scattered, often few and very large. It has quite a different appearance from the usual granular gland-cell. There are also present gland-cells of another type, which absorb eosin with avidity, and become dull brownish-red; these contain no black granules, and most often do not stain with haematoxylin. The ectoderm of the body- wall is much lower than that of the tentacles, but is well and evenly developed; it has scattered nuclei, and thick-walled enidae (shorter and proportionately broader than those in the tentacles) are plentiful, especially at the margin, where they are crowded. Gland-cells of a type intermediate between those of the tentacles and the common sort are found, but are not conspicuous; there are also eosinophilous ones. (b) Endoderm. The mesenterial filament-trefoils have rather narrow stems to the three lobes, and a good deal of mesogloea in each lobe, The ciliated lobes are often folded, Srepuunson—Cerluin Actiniaria collected off [reland. 151 and have nuclei throughout, but especially in their inner parts. The glandular parts of the filament have more ordinary granular gland-cells than anything else, and also a few non-granular ones. They contain thick-walled cnidae, irregularly distributed, variable in size and form, but typically fairly short, broad, boat-shaped, and blunt at the ends. Here and there a spirocyst (indigenous ?). (v) The above-described specimen presents no essential feature which warrants its separation from Verrill’s A. saginatus; the only difference is that his specimen had a mud-clasping base, but, as I shall show under Actinauge richardi, that is a non-constant character. My specimen does not widely differ externally from 4. nobilis, but that seems to have a definitely lobed disc, and the colour of saginatus agrees better with my specimen than does that of nobilis. 12. A, aurelia n. sp. “ Aurelia” from its general external resemblance to a jellyfish. (PI. XV, fig. 1; Pl. XVI, figs. 21, 28; Pl. XVII, figs. 10-13, 15, 16, and 18; Pl. XVILL, fig. 1.) 1. S.R. 335. May 12,1906. Lat. N. 51° 12’ 30’-51° 17’ 30”; Long. W. 12° 18—-12° 16’. 893-673 fathoms. ‘T'rawl. 6 specimens. 2. S.R. 335. (See 1.) 4 specimens. 3. S.R. 497, 499. Sept. 10,1907. 775-795 fathoms. Trawl. S.R. 497. Lat. N. 51° 2’; Long. W. 11° 36’. SR. 499. Lat. N. 50° 55’; Long. W. 11° 29’. 666-778 fathoms. Sept. 11, 1907. Measurements—(i) A large specimen. Diameter of oral dise, 10 cm. ; thick- ness of body-wall in a thick place, 1-3 cm. (ii) Smallest specimen. Diameter of oral disc, 7 cm. External characters—The entire facies of this animal is so unusual, that I found it difficult, at first, to believe that it really was a sea-anemone, and not a Seyphozoan. It may be thought that an animal so distinct in aspect from most other Actiniaria should be placed in a genus apart; and at one time I was inclined to think so. But comparison of the specimens with A. saginatus, and the fact that in all essential anatomical characters they resemble that species, have convinced me that it would be unwise to separate them. : The pedal disc is present in all cases, but is extremely reduced and insignificant; it forms a little pit or hollow with inturned edges, whose mouth is often completely closed. (PI. XV, fig. 1.) It contains traces of [S 2] 132 Proceedings of the Royul Irish Academy. dirt and mucus, and is sometimes radially striate. In the case in which it is best developed, its diameter and depth amount only to about 1-7em. It is quite thin-walled, thus differmg completely from the body-wall. The body is practically shapeless, its form when best preserved being that of a wide dish with a short thick stem. In all cases the dish-like part is bi-lobed, the two halves folding up against one another rather like the two valves of a lamelli- branch. Reference to the half of an animal which is represented in Pl. XV, figs. 1 and 2, will make it clear that the two halves of the animal are flat— fig. 1, as seen from the side (fig. 2), becomes much narrower. The thickness of the body-wall is rather uneven, but in the most regular cases it increases gradually from the edge of the pedal disc upwards, the maximum thickness being reached about the margin; and this latter is so thick and unwieldy that it rolls inwards over the tentacles and almost hides them. In one case the mesogloea here was 13cm. in thickness; it is soft and jelly-like and whitish. Margin continuous with tentacle bases. No verrucae, capitular ridges, ete. Body-surface more or less wrinkled, sometimes torn and damaged. The flat oral dise is very wide, and since it is so large and the tentacles are reduced, the animal, when opened out, looks like the under-side ofa mushroom. Disc, tentacles, and actinopharynx are deep purplish-black. Dise thin-walled, mouth fairly large, not widely gaping; lips prominent. Actinopharynx ridged. Tentacles entirely marginal, very small for the size of the animal—how they can be of any particular use to it, is difficult to imagine.. They are short, slender, soft, tapering, acuminate. Mach one is connected with the margin of the body by a thick, white bridge or lump of mesogloea, which is developed on its aboral side at the base. The tentacles are arranged regularly in ¢wo cycles only, the bridges of the outer cycle being usually better developed than those of the inner. The size of the tentacles and the development of the bridges vary to some extent in different parts of the rather uneven disc, and in most specimens some tentacles are missing— perhaps torn off. But the inner tentacles are submarginal, and seem to be on a level, so that one can trace no complicated arrangement such as is found in Porponia, ete. In three specimens whose tentacles I counted, the respective numbers were about 225, 260, and 275—that is, the total numbers. The disc is not many-lobed as in A. nobilis, but its curious way of folding seems to be a permanent character in consequence of which it is bi-lobed. Structure.—(i) Mesenteries. These are arranged in six cycles, the sixth not fully formed ; here and there one finds a pair even of a seventh cycle. Only the first four cycles are at all large, and only the six primary pairs are perfect, two of them being apparently directives. In consequence of the enlargement of the oral, and reduction of the pedal disc, and the general flattening of the Srepupnson— Certain Actiniaria collected off Ireland. 133 body, the coelenteron is a good deal restricted. The primary mesenteries are sterile; mesenteries of cycles 2, 2, 4 are fertile; those of 5 are_some of them, those of 6 are rarely, fertile. Two actinopharyngeal grooves. Mesenteries of all cycles bear filaments. I can see no marginal stomata in the perfect mesen- teries, but small oral-stomata are present. In the young cycles, each pair for the most part has one mesentery better developed than its partner. In the majority of cases this inequality follows the rule which prevails in Oa Sing giles vali Nit ae! neon Je) i “vi «hy ene seer ni uiieg eon) alae i Sadan tad) uy NAY cana s° il eu XT roon bad bee itn ae ai sil iaalt Peat 11 peter tiiepe +4 ig oor aa i ut @ he esa aaa a i : : 7 A. | hans i 1. ap iv wie as y webatadlts uP oy » ten wits ai Pu ie ‘ea? i a i : » wee 7 . i > € p! . ‘ Gye } nf) ad ip ’ aids! ' ’ el. news ee | ‘ r) ¥ we i yay TF {isa 1o% Jes. _ ' Moy | OU ae ' ¥ 4 » ¢ p AB OUTTOR Gs ou, May. Pla Al beh Og) ermine eal u ® H ‘ a4 : is 7 « 4 ” PROCEEDINGS OF THE ROYAL IRISH ACADEMY VOLUME XXXIV SECTION U.-ARUHBOLOGY, LINGUISTIC, AND LITERATURE. DUBLIN: HODGES, FIGGIS, & CO., LTD. LONDON: WILLIAMS & NORGATE 1917-1919 THE ACADEMY desires it to be understood that they are not answerable for any opinion, representation of facts, or train of reasoning that may appear in any of the following Papers. The Authors of the several Essays are alone responsible for their contents. Dust: Patrten at tHe Universtry Press ny Ponsonzy anp Gruns. CONTENTS SECTION €.—ARCH OLOGY, LINGUISTIC, AND LITERATURE, Armstrone (H.C. R.), F.S.A. :— Associated Finds of Irish Neolithic Celts, Armstrone (H.C. R.), F.S.A., and Rev, H. J. Lawnor, D.D., Lirr.D. :— The Domnach Airgid. (Plates III-V), c . Hoursy (J. R. D.), M.D. :-- See under Macauster (R. A.S Lawtor (Rev. H. J.), D.D., Lirr.D. :— See under Armstrone (H.C. R.). MacauistEr (R. A. §.), Lirr.D., F.S.A. :— Tamair Breg: a Study of the Remains and ‘Traditions of Tara. (Plates VII-X), Notes on some Ogham Inscriptions, ineluding fir recently “aise ered, Macauister (R. A. §.), Lirr.D., F.S.A., and J. R. D. Hoursy, M.D. :— Some Interments at Mooretown, Co. Meath. (Plate II), Manarry (Rev. J. P.), D.D., C.V.0. :— The Post-assaying found on dated pieces of Plate in the collection of Trinity College, Dublin, Power (Rev. P.) :— Place-names and Antiquities of S.E. Cork. I, Il, Westrorp (M. 8. D.) :— The Discrepancies between the Dates of Gift and Hall-mark Dates found on pieces of Plate belonging to Trinity College, Dublin, Westroprp (T. J.), M.A. :— The Ancient Sanctuaries of Knockainey and pipe: Co. Limerick, and their Goddesses. (Plate I). . The Karthworks, Traditions, and Gods of South-eastern Co. Tiineriel, especially from Knocklong to Tamair Erann. (Plate VI), PAGE 81 96 231 ~ 400 68 33 72 PP PPP PPP PP YP ERRATA. SECTION C. 59, 1.1. For Doonakemna read Doonakenna. . 149, 1. 24. For Retaliation read Retaliator. . 164, 1.24. For Donn read Dana. 165, L 11.. For bride of Macconmara read bride daughter of Macconmara. 165, 1.25. For Nuada read+Nuadat. 168, 1.3. Delete (d). 179, note 4. For N.W. read N.E. . 236, 1.10. For Mairisend read Mairisen, and so throughout the paper. 256, last line. For ater tone read later stone. . $31, L 1. For Dionysos, Zagreus read Dionysos Zagreus. . 350, 1. 4 from bottom. For “ bull-roarer’’ an oar, read bull-roarer “‘an oar’’. . 360, L 4 from bottom. For so read as. PROCEEDINGS THE ROYAL IRISH ACADEMY PAPERS READ BEFORE THE ACADEMY Ie PLACE-NAMES AND ANTIQUITIES OF §.E. CORK. By THE REV, PATRICK POWER. Read Decemprr 11, 1916; Published May 11, 1917. Purpose and scope of the survey hereby initiated will bear, and may require, a few words of introductory explanation. This work is the outcome, in part, of a proposal made, and practically advocated, some time since, by the President of University College, Cork, for a full and systematic study of the archeology of Cork county. In connexion with the study in question the present writer undertook—as his portion of the task—to collect the local names, and to note the antiquities of the south-eastern portion of the county. The region scheduled embraces the three baronies of Barrymore, Imokilly, and Kinnataloon, and was apparently, some of it at least, at one period, portion of the Decies.'| At the Synod of Rathbreasail the western boundary of the Diocese of Lismore was fixed at Cork, whence we may infer that in the early twelfth century the Deisi had extended their territory in that direction. The task involved in the proposed survey is by no means easy or light; it is heavier far than the casual reader dreams, and, in the ordinary course of things, it will take some years to accomplish it. It postulates a personal visit, not only to every townland of the wide area described, but, in nearly every case, to several parts of each townland; it requires the interviewing and cross-examination of hundreds of Ivish speakers and other residents ; 1 Martyr. Gorman, p. 228 ; Martyr. Donegal, pp. xxxvii and 321 ; MacErlean, ‘* Synod of Raith Breasail,” p. 25, in Archiy. Hibernicum, vol. iii, &e. R.I,A. PROC., VOL, XNXIV., SECT. C. [1] 2 Proceedings of the Royal Irish Academy. it demands, finally, measurements and critical examination of historic and pre-historic monuments, as well as much collation of ancient documents and authorities. For convenience’ sake and facility of subsequent reference, the writer proposes to follow the system of treatment by parish and townland, first adopted by him nine or ten years since, in his “ Place-Names of Decies.” Baronies, parishes, and townlands will be taken alphabetically—parishes under their respective baronies, and townlands under their respective parishes. Under townlands will come townland sub-denominations, when there are any. In enumeration or record of sub-denominations some dis- crimination is necessary. There are thousands of sub-denominations not worthy of record; these are such frequently occurring names as “ Big-Field,” “Lios (or Fort) Field,” &c., which may be regarded as merely generic or descriptive, and really common nouns or adjectives. Our place-names naturally arrange themselves in strata, according to age, somewhat after the manner of geological deposits. Some of them are of the remotest antiquity; others are of the last generation; between the two extreme layers some three or four others are more or less clearly defined. The majority of townland names are of great age, while most of the sub- denominations are comparatively recent. River, mountain, and territorial names seem the oldest of all. Nearest the surface, i.e. most recent, are, as a rule, names of fields, wells, and roads. A fair proportion of sub-denomina- tions are, however, undoubtedly old. Of the townland names probably nearly fifty per cent. go back to the thirteenth century. Some of them indeed are older than history, and date from pagan times. Of all these various strata, and of some intermediate classes not enumerated, our south-east of Co. Cork affords copious illustrations. Generally speaking, mountain and maritime districts yield most (though not necessarily the most interesting) names, and demesnes and urban areas the fewest. The present unit of Irish land division is the townland—a denomination now of very varied extent and character. Strictly speaking, the area of a townland is, or should be, 480 acres. The ancient Irish survey was by rough computation, and not by chain; moreover, bog or mountain was not counted. At the time of the Ordnance Survey, however, many smaller divisions or denominations were raised to townland rank, and some amalga- mation of townlands was effected; hence, we have to-day townlands varying in area from an acre to thousands of acres. The chief ancient Irish land sub-divisions were, in ascending order, the gneeve (Gniomh), sesseagh, tate or ballybo, seisreach or ploughland (also called carrow, carucate, or quarter), the ballybetagh or townland, and the triocha-cead or barony. All these terms, Powrr—Place-Names and Antiquities of SB. Cork. 5) or most of them, we shall find incorporated in the place-names of the region with which we deal. A gneeve (Gniomh) measures, roughly, ten acres; a sesseagh is equivalent to two gneeves; a ballybo (Bailebé) is thrice as extensive again, while a seisreach or carrow (Ceathramha) equals 120 acres, and a ballybetagh (Baile-biadhtaigh) contains four seisreachs. The Tricha- Cét (Triocha-cead), which became the modern barony, contained thirty ballybetaghs.!. According to Keating the Baile contained twelve Seisreacha. The truth seems to be that the Baile (and, proportionately, the Seisreach) varied somewhat in area, according to the period and the population, as indeed also did the acre, according to the stock it could carry. 4 Up to the forties of the last century some of the lesser sub- denominations were recognized as official sub-divisions, but, since the Ordnance Survey, all divisions below townlands have been discarded, and the term “townland” is now applied to every official sub-denomination, whatever its area. The parish is, in its origin, a purely ecclesiastical divi- sion. In course of time, however, it came to be recognized for various civil purposes. Generally speaking, the parish name is taken irom the townland in which stood the ancient parish church. Sometimes, however, the church (e.g. Kilshanahan, Killaspugmullane, Templeusque) gives name to the parish. Needless to add, the parishes of the present paper are not the modern, but the ancient, ecclesiastical divisions, now often styled civil parishes. These are, from the point of view of the local historian and the antiquary, the most important divisions of all. Unfortunately the new Ordnance maps omit them entirely. In collection and interpretation of the place-names here recorded, my primary care has been to get the name from the lips of a local speaker cf Trish. Observance of this precaution is a matter of vital importance, for nothing, except the authority of an ancient document, can replace the native, $ traditional pronunciation. To negiect of it most of the “howlers” in popular place-name etymology are traceable. A fact which it behoves the student of this subject to keep well before his mind is that the place-name is nearly always a very simple thing. When first applied the name was intended as descriptive merely. Hence, in its invention, there was no conscious striving at effect, and there is much less poetry latent in our local names than irresponsible local histories and fatuous guide-books would have us believe. ‘There are, of course, occasional flashes of fancy, but they are based, as a rule, on resemblances, more or less apparent, to some object or animal. ‘ Harding, ‘On MS. Mapped Townland Surveys of Ireland,” Proceedings R.I A., vol. viii, pt. 1; Windele MSS., R.I-A.; MacNeill, ‘‘ Barly Irish Population Groups,” Proceedings R.I.A., vol. xxix, pp. 102, &e. (1*) 4 Proceedings of the Royal Irish Academy. ABBREVIATIONS USED. A.F.M. Annals of the Four Masters. A.S.E. Acts of Settlement and Explanation (seventeenth century). B.S.D. Books of Survey and Distribution (seventeenth century). DS. Down Survey (maps and references). Ing. Inquisition. O.M. Six-Inch Ordnance Map. 8.DD. Sub-denominations of Townland. Sub-div. Sub-division. Tax. Papal Taxation (twelfth century). ~ Visit. Ecclesiastical Visitation Book (end of sixteenth or beginning of seventeenth century). BARONY OF BARRYMORE. The barony derives its name primarily from the Anglo-Norman De Barrys, whose territory it became. More particularly, however, the name came from the Lords Barrymore, whose barony the region bearing it was constituted at an early period. The ancient designation was Ui (Uibh) Leathain, Le. the tribal land of O’Lehane, a name perpetuated in Castle Lyons. Ui Lehane probably embraced Ui MacCaille too.' Barrymore is a barony of large extent, embracing twenty-nine parishes and parts of parishes, and covering roughly some 153,000 acres. Included, besides the great and little islands, are the islands of Spike, Hawlbowline, Foaty, and Harper's, &c., in Cork Harbour. Fermoy and Condons’ baronies bound it on the north; Barretts’, Cork, and Kerryeurrihy on the west; Cork Harbour and Imokilly on the south, and Kinnatalloon and Imokilly on the east. Physically, Barrymore is undulating, or, rather, decidedly hilly, in character, but there is little actual mountain. The soil is of at least average fertility, and the condition of the people is comfortable in the main. Through the barony run the main Cork-Fermoy and Cork-Youghal roads, and this fact, coupled with the situation of towns like Cork, Midleton and Fermoy, practically on its borders, explains to some extent the considerable Anglicization that has taken place. Irish is hardly spoken at all at the present time, and in some parishes I found great difficulty in discovering an Irish speaker who could remember the tra- ditional pronunciation of the names. Fortunately, however, desuetude of the native tongue is so recent that much of the folk-lore is still recoverable. The picturesque and homely thatched homesteads of the nineteenth and ' Book of Leinster, p. 360 ; Martyr. Donegal, p. 129. Powrr— Place-Names and Antiquities of SE. Cork. 5 eighteenth centuries are fast disappearing or have disappeared in Barrymore, and ugly two-story slated dwellings are, thanks presumably to Board of Works plans, taking their place. Perhaps it is not quite fair to associate the ugly, modern houses with destruction of ancient monuments, but it seems a fact that where the new houses are most common, vandalism has been most rampant. Within the whole barony I scarcely remember to have seen a quern-stone, and I can recall but one dresser of pewter. In the neighbouring territory of Decies both the quern-stone and the set of pewter- plates are quite common, though neither quern nor pewter has been in use for half a century; the objects are preserved for old family associations. Similarly, in Barrymore, I have found scarcely any trace of the Béyha Bride (or custom of plaiting a cross of straw on St. Brigid’s Eve) or of other folk-customs prevalent in the Decies, The tally-stick, for account of labourers’ wages, is not in use, though the older people remember it, and it still survives in other parts of Co. Cork. Lioses, many of them fine specimens, and in excellent preservation, abound, though hundreds have been levelled in recent years. Pillar-stones are also very numerous; so are holy wells, and devotion to the latter is practised throughout the whole barony. An inexplicable phenomenon is the almost total disappearance of ancient church remains. Of the twenty-nine parish churches, and the many connected chapels of post-invasion times, not more, perhaps, than a dozen in all have left any remains; the others have been literally razed to their foundations. This wholesale clearance, as well as some of the other phenomena, or absence thereof, referred to already, I feel inclined prima facie to attribute largely to the Desmond war and its consequences. Strange that the Cromwellian plantation of the neighbouring counties should have produced results so very much less marked. PARISH OF ARDNAGEEHY. Ardnageehy parish, which, on its north side, runs well into the Nagles Mountains. includes a considerable area of wild and rugged country. Con- sequently its local names are of somewhat more than ordinary interest and number. Many of the older people—though few of them speak irish freely— retain sufficient knowledge of the vernacular to be able to quote the correct forms of the names. The parish is called, in the usual way, from the town- land in which the ancient church and graveyard are situated. An older name was Garthenegaythe,! which may be rendered—* Breezy Garden.” Of the pre-Reformation church, which was a plain oblong in plan—twenty 1 Tax., 1291. 6 Proceedings of the Royal Irish Academy. yards in length, by seven yards wide—only the west gable and the founda- tions of the north side-wall survive. In the gable is a single, small, oblong ope, which splays inwardly, and was intended evidently to light a chamber or gallery, partly supported upon a projecting ledge at the west end of the interior." Above the ledge referred to, the thickness of the gable is about three feet, as against a thickness of four feet from the ledge downwards. There is a small, nearly square holy-water stoup, or font, measuring roughly some eighteen inches at the sides. The surrounding graveyard—some half- an-acre in extent—possesses harcly any feature or monument of interest. The earliest inscription I found was dated 1770. I was unable, unfortunately, to ascertain date of the “ pattern.” TOWNLANDS, Arparow, ‘Ard a Rogha.—* Chosen (or Choice) Height.” Area, 369a. This name, at least as a townland designation, seems to be of comparatively recent origin. Formerly the place, which is mostly mountain, was regarded as a sub-division of Lackendarragh. Mr. R. A. Foley suggests “Ard Idir dé Abha.—* The ‘tween Rivers’ Height,” instead of ‘Ard a Rogha, and very likely his interpretation is correct. S.DD. An Branar—“ The Fallow Field.” Tobar an Uisge—* (Household) Water Well.” Tobar na Leacht—“ Well of (beside) the Monumental Piles.” The piles or heaps in question are of stones, and apparently of natural origin— simulating, however, the artificial. I am not at all sure of this name. First, I doubt that the term “leacht” would be applied in this way to a natural stone-heap. Secondly, my informant, who was an indifferent Irishian, and not overburdened with intelligence, wavered somewhat between leacht, lag, and leac. He inclined most to the first, and I could not, on the occasion, get his testimony corroborated. I imagine the word is really leacan, a glen slope. “The Corragauns ” (Na Carragéin) —* The Little Rock-abounding Fields.” An Fhaithche—“ The Lawn (Green, or Hurling Field).” An Seagal—“ The Rye (Field).” “The Sweeteen” — A field deriving its name from its succulent pasturage. Seana Garraidhe—* Old Garden.” ! This gallery was quite a common feature in churches of the present class. It may have been intended for the ecclesiastic’s lodging. Sometimes, instead of the gallery, we find at the west end a chamber entered from within the church, and occasionally we have traces of both gallery and chamber. Powrr— Place-Names and Antiquities of S.E. Cork. 7 Paire na dTor—* Field of the Bushes.” Paire na Mona—< Turf Field.” ARDNAGEEHY, ‘Ard na Gaoithe—*‘ Wind-swept Height.” Area (in two parts), 10754. The name-giving “height” is, presumably, the exposed elevation half-a-mile, or less, to south-west of the ruined church. The church itself is on comparatively low and level ground. Other antiquities of the townland are four lioses (O.M.), and a small pillar-stone. The last- mentioned object is on the farm of John Sheehan. One of the four lioses, situated on west side of the townland, has been levelled, and its souterrains have been partly destroyed. Ardnageehy (Ing. Car. I). $.DD. Crosara na Con—* The Dog’s Cross-roads.” The eponymous canine was an enchanted hound, or a woman who sometimes took the shape of a dog or sheep. This dread creature frequented the present cross-roads at night, injuring and sometimes killing inoffensive wayfarers. A former parish priest of the district met his death at the place, but a successor of his, a Father Falvey (?), is popularly believed to have exorcised and finally banished the malicious and supernatural visitant. Exorcism and banishment notwith- standing, the spot is still dreaded, and daring is the man who would pass there alone after nightfall. A cattle-fair was formerly held at Crosara-na-Con. ‘Ard Maganna—Meaning unknown ; a sub-division of no great area ; also an old name for the whole townland, according to Denis Dunlea, an intelligent resident. Réidh a Gheilte—* The Lunatic’s Mountain-Plain”; a former wild spot of some six or eight acres, now cultivated. Bun a Bhaile—“ The Homestead Hollow.” Paire na Ceardchan—* Field of (adjoining) the Smithy.” Paire na Noinini— Daisy Field.” Paire a Leasa—“ Lios Field.” Paire na bhFear—“ The Men’s (Meeting) Field,” i.e. the place where men assembled on Sundays or on summer evenings. Paire na mBréithre—“ Field of the Judgment (Verdict)”; evidently its ownership was the object of a lawsuit. Pairc na mBullana—* Field of the Large Round Stones (Bowls).” 1T translate Baile uniformly as homestead, though I feel that this word does not always express the exact sense of the Irish term. The ordinary rendering, town, is less exact and somewhat misleading. An alternative rendering, stead, has been suggested ; but I do not think that it has any advantage over homestead, and it is not as well understood. In Australia, the word ‘“‘ homestead” is in constant use to signify the same thing exactly as our Irish Laie, 8 _ Proceedings of the Royal Trish Academy. Garraidhe Fada—* Long Garden”; locally, by garden is meant a tillage- field. Faithche—* Paddock (or Hurling Field ov Fair Green).” BALLINLEGANE, Baile an Liagain—“ Homestead of (beside) the Pillar- Stone.” Area, 631 a. Dallilega (Depositions, 1652). The word baile, by the way, enters into the composition of no fewer than 6,400 Irish townland names.'' Names incorporating the term we may regard as almost exclusively of post-invasion origin. The eponymous pillar-stone, which stands in Twomey’s farm, is a massive, and indeed magnificent, example of this class of monument; it is almost square in horizontal section, and measures nearly twelve feet in height by from thirty-eight to forty-five inches at the sides. Close to the great dallan, and in the same large field, is a well- preserved circular lios of medium area, with rampart some six or eight feet high. On the townland are two further lioses, also in fair preservation. Near the south-west angle of the townland, on the farm of Mr. Cashman, is a second, but much smaller, standing stone—also, apparently, a true dallan. S.DD. Seana Bhaile—*Old Homestead (or Village)”; a sub-division,, consisting of three or more large fields, on which a fair was formerly held. “The Faheens” (Faithchinidhe- —‘The Paddocks (Greens or Hurling Fields).” Paire na Cloiche —“ Field of the Pillar-stone; 2 from the great dallan above referred to. Ban a Bhaile—“ Home Field” ; it adjoins “ The Faheens” above. Tobar a Tae—* Tea Well” ; now drained by a subterranean aqueduct. “The Young Fields” ; two large fields. ‘The Scorched Field,” “*'Uhe Lumpy Field,’’ and “ The Lime Field.” Bisnor’s IsLanp, Caislean an Easpoig—‘“The Bishop’s Castle.” Area, 587 A. The Ordnance Survey field books give Oile4n an Easpoig as the Irish form ; but I distinctly got Caisléan from the only local speaker of Irish (an intelligent and trustworthy man) who had ever heard an Irish form. Perhaps Caisledn is a mistake for Oileén, or vice versa; the two words could be easily econfused.* The name Ballinaspuigmore occurs in the Depositions of 1652, ' Reeves, ‘“‘ Townland Distribution of Ireland.” * In local usage cloch is generally a pillar-stone, with which dallan, gallan, liagan and cloch fhada are synonymous. ® Oiledn is one of the words of occasional occurrence in Irish place-names, the true secondary sense of which it is difficult to fix. Often, of course, its force is our English “island” ; but there are many cases, like the present, in which the literal sense is impossible. It has been suggested that the toponomical oiledn is a small area surrounded Powrr— Place-Names and Antiquities of 8. B. Cork. 9 and most likely it designates the present townland. Consequent on agri- cultural improvements (7), a fairly clean sweep has been made of the former antiquities ; the only items surviving are—one large circular lios, almost perfect, and a solitary pillar-stone, five feet high by three feet six inches and one foot nine inches—both on Mr. James Cashman’s farm. On surface of the pillar-stone are some natural cup-hollows. Built into a gate-pier, at entrance to the field in which stands the pillar-stone, I found an ogham-inseribed slab which I had removed to University College, Cork.!' The inscription which, unfortunately, is incomplete, reads: “Olagni Maqi....” (Of Olagnos, the Son of....). Sub-denominations are very few, and portion of the townland lies within the neighbouring parish of Kilquane. S.DD. Paire ‘Ard—“ High Field.” Bogach ‘Ard—“ Elevated Swamp.” The name is applied to a field now quite dry and arable, but sporadically producing bog-plants which testify, as does the name, to its former marshy character. BrIDESOWN Baile na Bride. ats, Cnoc an Chatha—‘“ Homestead by the Bird’s River,” otherwise “Hill of the Battle.” Area, 9054. I got the first name from P. Murphy, Edmondstown, and the second from P. Kennedy, who could not say for certain whether it was a synonym for the first or merely designative of a sub-division of the latter. I failed to find local corroboration of the second name ; the first was amply corroborated, although the Ordnance Surveyors do not record any Ivish form. Reasoning from failure of the Surveyors to find (or record) an Irish form, it looks as if the present official name is modern; moreover, we do not find it in the usual seventeenth-century documents. From its appearance, however, I am inclined to think it some centuries old. S.DD. “Bride River” (O.M.)—Bride; the name of a Celtic goddess, from root Bri, strength, &e. Compare Brigantes, the name of a well-known Celtic tribe. Our present Bride, which joins the Blackwater, is sometimes styled Bride Barracha to distinguish it from the B. Musceraidhe, a tributary of the Lee. Cuile—“ Back Place,” the name (locative case)’ of a well-known sub- division. Cul, gen. cuil, and cuil, gen. ctile, are very difficult, and often impossible, to distinguish in place-names. Of course one is masculine and by a stream or marsh ; even this does not explain away the difficulty, for the oiledn is not always so surrounded. Perhaps the word had reference to proprietorship or estate ; the place may have been a small property surrounded on all sides by another estate. 1 See ‘‘Ivernian Journal,’”’ vol. vi, pp. 201, &e. 2 Use of the dative (locative) for the nominative is fairly frequent in Irish place- names. The usage doubtless arose from association of the name with the idea of motion thither or thence. R.I.A. PROC., VOL. XXXIV., SECT. OC. [2] 10 Proceedings of the Royal Irish Academy. the other feminine; but the gender is of little help, simce both words generally appear at the beginning of the name. It is some comfort in the matter that the words do not differ greatly in meaning. Poll Lughaidh—* Drowning Place (Pool) of Louis”; a deep pool in the Bride, the name of which has been corrupted to Poll-a-Wig, in popular usage. Poillin na bhF6d—* Little River-Hole of the (Turf) Sods,” in allusion probably to the use of burning turf in poaching. “The Turning Hole,” a third river-hole and small whirlpool well known to anglers and poachers. Céim—* Pass”; a sub-div. containing some twenty acres. CyoppocE, Cnapog—* Knoll.” Area, 550 A. This townland is almost entirely uninhabited, although as long ago as the year before the famine twenty-five acres of it were under cultivation. $.D. Seefin (O.M.) Suidhe Finn—“The Sitting Place of Fionn (Mac Cumhaill)”; a name of frequent application to mountain peaks and cairns. The present Swidhe is at the extreme north-east angle of the townland, and at an elevation of 1392 feet. BUNNAGLANNA, Bun a Ghleanna—“ Glen Bottom (End).” Area, 247 a. The old people use Bun a Bhaile, “ Village End,” as a synonym. Agricultural occupation is recent, and that the name is comparatively modern is suggested by its absence from ancient documents. S.DD. Piast an Uisge—* Water Serpent”; I give the name as I got it, though I cannot explain its application here to a field. Poll an Rudaire—‘ The Knight’s Pool”; a hole in the Bride River. perhaps the Knight’s drowning place. Connlach Liath (“ Grey Stubble”), Currach (“Swamp”), Barra Leacan (“Summit of the Glen-Slope”), Fiadhnan (“ Wild Place”), M6éin Loman (“Bare Mountain Bog”), and Cnoc4n Ruadh (‘Little Red Hill”)—the names of a number of fields. Carric, Carraig—“ Rock.” Area, 325 A.; a long and narrow strip of mountain land, thinly populated, and in part uninhabited. $S.DD.—Two fields, named respectively Seana Ghdirdin and Seana Phéire— “ Old Garden ” and “ Old Field.” CHIMNEYFIELD, Paire a tSimné—* Field of the Chimney.” Area, 141 a. The name is derived from a ruined house with a standing chimney, once almost the only artificial feature on the townland. Reclamation has been recent. Ath a Dina appears to have been the older name. The Dun is not necessarily a rath or lios, or any artificial structure at all; sometimes, especially in glens and mountain regions, it is a natural mound, simulating PowER—Place- Names and Antiquities of S.L. Cork. 11 a rath in its rounded and regular appearance. At the date of the Ordnance Survey one Catherine Heaphy was engaged in teaching Irish here; there was also, at the same period, a private school attended by about forty pupils, who paid 2s. each per quarter for tuition in reading; 2s. 6d. for tuition in reading and writing ; and 3s. for the whole three R’s.! S.D. Carraig an Aifrinn—“ Mass Rock”; a natural outcrop or cliff by the river, in the shelter of which, presumably, people met feloniously to worship in the Penal Days of saddening memories. Commons, Coimin and Réidh a Choimin.—Idem. Area, 389 A.; all mountain, and practically uninhabited. The Ordnance Surveyors? add a note—“John Hide, Esq.,and Pierce Nagle, Esq., each claim the whole as their property. None of it cultivated, but very capable of being so. No rent paid for it.” S.DD. An Leacht—* The Monumental Pile,” on the mountain summit. Leachts deserve to be classed as a special type of rude stone monument ; they vary enormously in age—from the Bronze period to the nineteenth century. Even within the past sixty years leachts have been raised to commemorate tragic events like murders and deaths by accident. Leachts —like our present specimen—on mountain summits are of the oldest variety, and date from prehistoric times. Féith Ghorm. “Bluish-Green (Dark) Vein.” “ Féith,” of rather frequent occurrence in mountain and bog names, is applied to the luxuriant green band or patch which marks the course of a subterranean spring. In the present instance the Féith is a well; originally the name must have been applied to the strip of coloured herbage which indicated the course of the spring issuing from or feeding the well. CorBALLY, Corra Bhaile—“ Round Hill of the Homestead.” Area, 451 A. This isa townland name of fairly frequent occurrence throughout Ireland. O’Donovan generally, if not universally,? explains it as “Odd Town,” Le. Cor Bhaile, which I venture to designate incorrect. Joyce follows O’ Donovan, and I may appear guilty of temerity in differing from such eminent authority, but I have always heard the name pronounced Corr a Bhaile. Corbally (D. 8S. Reference). S.D. Fleise—* Wet Place”; a sub-div. of some hundred acres. There is also a Flesk River in this barony, as well as one in Kerry and another in Antrim. GLENNASACK, Gleann na Sac.—‘Glenn of the Sacks”; presumably ' Ordnance Survey Field Books. 2 Field Books, Ordnance Survey, Mountjoy Barracks. “ Field Books, Ordnance Survey. 12 Proceedings of the Royal Irish Academy. because, owing to want of roads, loads had to be carried in sacks slung across horses’ backs. Area, 485 A., largely mountain. S.DD. An Réidh—* The Mountain Plain”; a field. “The Dummy Field”; a large semi-reclaimed enclosure of perhaps forty acres, which is said to have been “ graffed,’ many years ago, by two or three deaf mutes. GLENVILLE, Gleann a Phreachdin—“Glen of the Crow”; originally, according to one authority, the name was Gleann na bhFraochan (“ Walley of the Whortle-berries ”). Area, 941 4. SDD. An Biolarach (The Biolarach)—“The Watercress-abounding Spring.” This is the same as the Irish name of Watergrasshill; it is applied in the present instance to a well. “ Coalbog Wood ” (O.M.), from a superior kind of hand turf made there, and said to resemble sea-coal in quality. Cumar na Sagart—“The Priests’ River Confluence”; perhaps because the priests’ residence was near by, or because the confluence marked the meeting-place of coterminous parishes, or again, perhaps because some priests were drowned in the river at this spot. Gleann Ban—* White Glen.” Caisledinin —“ Little Castle”; the name is now applied to a stream, but originally it designated a rock which overlooks, or overhangs, the river. “ The Stone Field,” within Sir E. Kinahan’s demesne. Here once stood a dallan “ as large as the Ballylegane specimen.” Throughout the Barony the “stone” or “cloch” of the numerous “Stone Fields”’ and Paircs-na-Cloiche (or P. na-gCloch) are really pillar-stones. Alas, that so many of them have disappeared ! Bun a Bhaile and Barr a Bhaile—“ Village End” and “ Village Top” respectively. “ The Tinker’s Hole,” a river pool in which, presumably, the strolling man of solder met his untimely end. “The Weir Inches "—Inse (Inch) is the genitive case of Inis, an island, and the word is commonly applied to a river-holm or a low-lying meadow by a stream. “Glashanabrack River” (O.M.). Glaise na mBreac—“ Trout Stream,” on the east boundary of the townland) The common name now is “ Trout- brook,” in which an effort is made to preserve both the sound and the sense of the original. Graicur, An Ghraig—* The Village.” Area, in two parts, 1,161 4. Graig—Ing. Car. 1. S.D. “The Bow Field” 1 met nobody who could throw any light on Power—Place-Names and Antiquities of SL. Cork. 13 origin of the name. Perhaps it was a half-vendering of Paire na mbBo, “ Wield of the Cows,” or of P. na mBoth—*“ Field of the Tents.” INCHINANAGH, Inse na nEach.—‘ he Horses’ River Holm.” Area, 230 A. S.DD. Barr a Bhothair—* Top of the Road”; applied here to a cross- roads. Cluain—“ Meadow”; a field. “The Blacks”; applied to a group of small fields in allusion to the colour of their peaty soil. “The Long Streak” and “The Short Streak ” = two fields. S¢reah here is really the Irish, Straic, a portion of anything involving length. Poll a Choreéin —“ The Pot Hole” ; in the river. Kiiieacu, Coill Liath—“ Grey Wood.” Area 6314. “Torna” states that the name is Cill Liath—“ Grey Church.” I believe the townland is now uninhabited, though it was not so at the date of the Ordnance Survey, when it produced thirty barrels of potatoes and six of oats to the acre. Jilleagh is a very long and narrow townland. S.DD. Boigthin Ruadh—* Little Soft Red Bog”; a bog or sub-division of considerable area, near the mountain summit, which produced an excellent brown turf. Féith na Bearna— Vein of the Gap.” I éith has been already explained as the designation of a green, grassy line in a mountain bog indicating the course of a spring. KOLLUNTIN, Cill Fhionntain—“ Fintan’s Church.” Avea, in two divisions, 9144. Site of the ancient Celtic Church which gave its name to the town- land is now occupied by Killuntin House. A man who saw the old graveyard sixty or seventy years since describes it as surrounded, more Hibernico, by an earthen fence of the usual circular type. No remains now survive. S.DD. Moin na Snathaire—Meaning unknown; perhaps M. a Snamh- aire. It is the name of a sub-div. and the site of a former village. Paire na gCaorach—* Sheep Field.” “The Black.” (See under Inchinanagh, antea). KNOCKAUNLOUR, Cnocan na Lobhar—‘ The Lepers’ Little Hill.” Lepers may have lived there, or the lands may have been part of the endowment of an hospital for lepers. Knockaunacarren seems to have been an older name ; at any rate it was the name given to the Ordnance Surveyors by R. G. Champion, Esq. Area 545 a. S8.DD. Comerboy Bridge (0.M.) on south-east boundary. Cumar Buidhe —“Yellow River Confluence.” The Cnocans (Cnocdin—“ Little Hills’’); the fields surrounding a natural Dan, or gravel mound, on Ford’s farm. 14 Proceedings of the Royal Irish Academy. Paire an Athain— Little Ford Field.” Biolarach—(See under Glenville above); a well. Seana Phina—‘‘ Old Pound”; the name ofa field. The “ Pound” was more than a mere place of internment for straying animals or the sheriff’s seizures. It was also the temporary depository of the tithe-proctors’ prey ; and the word was used locally in the very wide sense of a safe enclosure or yard for cattle. Gort Dubh (“ Black Garden”), Paire a tSeagail (“ Rye Field”), Pdircin a Round or P. na Rann (?), Faithce (“ A Lawn”) and Féithin (“ Little Green Vein ’’)—a series of interesting field-names. Kyockapoorty, Cnoe Ui Dubhartaigh—“ O’Doorty’s Hill.” O'Donovan, who writes the qualifying term, Dirta, says the meaning is very uncertain. I, however, found the name as above. Area, 358 a. The townland is entirely uninhabited, and its name is almost forgotten in the locality. Knoppoce, Cnapog —* Hillock.” Area, 358 A.; almost uninhabited. LACKENDARKAGH, Leacan Daraclhi—‘ Glen-slope of the Oakwood.” Area, in three divisions, 2,022 A. Observe the use of the oblique Leacan for the nominative. There is one lios, in tolerable preservation, on Murphy’s farm. Lackendaragh (D.8. R.). S.DD. Céim Carraige—“ Rock Pass.” Béthairin Dearg—*“ Little Red Road”; on south side of townland. Tuar a Bhoth4in—“ Cattle Night-field of (with) the Shed.” Cairn—“Stone-pile” ; the name of a small sub-div. Clais Ui Chathail—* O’Cathal’s Trench.” Carraig an Aifrinn—‘* The Mass Rock.” “The Lawn” (Dall4n—A Pillar-stone); a mountain sub-div. of some fourteen acres. Paircin an Airgid—* Little Field of the Money”; in allusion to alleged buried treasure. Various Field Names :—Paéire na gCloch (“Field of the Stones”; most probably pillar-stones); Péire na Sgeiche (“Field of the White-thorn Bush”): Paire na gClir (“Field of the Pigeons”); Paire na gCoinin (“Field of the Rabbits”); Paire a Chumaraigh (“Field of the Hills and Hollows”’); The Pounds; The Blacks; Faoifaun (probably Faoi-Faénaidh) (“ Below the Incline”); Bandg (“Little Field” 7), &., &e. Lyravarric. Ladhar a Bharraigh—* Barry's River-fork”; so named from junction, at this place, of a small, nameless river with the Bride. Area, 495 a. $.D. Cnocan—* Little Hill” ; a sub-div. Powrr— Place-Names and Antiquities of S.E. Cork. 15 MEENANE, Minean—This is a mountain name of fairly frequent occurrence and signifies, according to O’Donovan,' a small green patch on a mountain. Area, 559 a. There was a lios, not noted on the O.M., in a field on Towhill’s farm. S.DD. “ Watergrasshill” (O.M.). Cnocin na Biolaraighe—(“ Little Hill of the Water Cress”’). Strictly speaking Biolarach is the stream which flows through the watercress, or in which the latter grows. It is not necessary to remind the reader how this place derives its fame from the former residence here of one who literally had greatness thrust upon him—Father Daniel Prout, “Of Waterevasshill, the renowned P.P.” The good priest died in 1830, and is buried at Ballinaltig, a couple of miles from the village for ever associated with his memory. A stream, which rises in Bishop’s Island, ran down the whole length of the village street, and in the old mail-coach days there was a pool for watering the horses. Water- eress flourished in the stream and catchment pool in question, and hence the name. Immediately previous to the famine, there were one hundred and thirty inhabited houses in Watergrasshill, and a population of some five hundred and twenty. Bothairin Bhlaic—“ Little Road of the Peaty-surfaced Place.” The word Blac, occurring occasionally in mountain-names, is really the English “black,” which is used as a noun to designate the dark soil of reclaimed mountain land. Béal a Chreathaigh—(“ River Mouth of the Hurdle Bridge”); a sub-div. Compare the Latin, crates. “« Blackstone Bridge”’; this place was a Mass-station in the Penal times. Ladhar na nGleann—“ Fork of the (two) Glens”; another sub-diy. Paéire a Leasa— Of names in the Thomond legends, found also on British and Gaulish monuments, we have Art, Adaros, Brigios, Camul, Cass, Coman, Cona; the prefix Cuno, Cathall, Dara, Garmanos, Lugus, Lugudos, Mor, Neden, Nem, Nemet, Segomo, and Vlattu. (See Rev. Celt., vol. i, pp. 293-8.) * Nuada, so described in Agallamh, p. 132. The estuary was, doubtless, his ‘‘ silver hand (most vivid picture of a distant river-mouth), as Boand’s “ silver fore-arm,” is the Boyne River. The epithet was probably invented in Britain, where the great river- mouths impressed the Gauls. We see ‘‘ Blood-red Nuada silver hand” in a poem of St. Columba (Ossianic Society, vol. v, p. 255). * * Lugh, like the sun in splendour, men were not able to look on his face” (Atlantis, vol. iv, p. 161). *“Segomar . . . made Bélisama (the crafts goddess) this nemeton”’ (Rhys, Hibbert Lectures, vol. iv, p. 46). See Anc. Irish Laws, vol. i, p. 165, for ‘‘ Fidh nemidh” and “* Defidh ” ; also burning of Fidh nemidh, Armagh, 4.p. 996. Drunemeton was a chief sanctuary in Galatia. The nemeta were scenes of human sacrifice (Lucan, Pharsalia, book m1). Taranis, Teutates, and Hesus were the Gaulish triad to whom human beings were offered. Perhaps the Irish equivalents of the first two, Etherun and Segomo, at Tara and in South Munster, were thus honoured. Cenn Cruach was perhaps a “‘ hill-god” (‘*head of the mountain”); he received human offerings ; and the summit of the Cruach of Croaghpatrick, in Co. Mayo, girt by a great dry-stone rampart, appears as the scene of a spiritual struggle of St. Patrick with the demons and with ‘‘ Crom Dubh.” Westrropp—Ancient Sanctuaries of Knockainey and Clogher. 55 were of the latter species. The grove of ‘Aine was clearly of the second “Trees, But who was ‘Aine? The occurrence of the name of Aife at Gleneefy, and the lost “Cenn Aife,” not far from ‘Aine’s hill, suggests that the two ladies may be the fairy goddesses ‘Aine and Aife,' who, with Clidna and Aibinn,? figure so largely in Munster folk-lore. The first two are closely connected with Manannan mac Lir,* the sea-god. The “Duanaire Finn” tells how the lady Aife was transformed into a crane in his house, and how he killed Fer Fidhail,‘ son of Eogabal, of Sid Cliath, who was his pupil, and bore a forked javelin (? trident). Another tale in the “ Agallamh”® tells how ‘Aine, daughter of Eogabal, married him (Manannan) on condition that he gave his wife to her brother Aillen.* Aife was wife of Cain, of Cenn Febrat slebe Cain, evidently one of the group of mountain-gods in Clfu.? Others said the ladies were daughters of the sea-god, and ‘Aine at least was a goddess of fishers ap Dunainey, on the coast.’ This does not make the Aillen tale impossible in the known laxity of pagan ideas,* for what was abysmal sin in the law of Sinai and Galilee was sacrosanct among the kings of Persia, Peru, and Egypt. In a late tale the sea-god’s son, Etar, died of love for ‘Aine, and was buried (a late mark) on the peak of Edar* or Howth—the “ Edros” of Ptolemy. In the “Sleep Song of Grainne,”’® over Diarmait, we hear of “ the 10’Rahilly’s Poems (Ir. Texts Soc., p. 203) place Aife’s sidh in Sliab Eibhlinn. “< Aife’s cows’’ were rocks on a hillside (see Rev. Celt., vol. xiii, p. 378; and Metr. Dind., vol. x, pp. 115, 231, and 499). Her name meant ‘‘hillside.” For Aoife, wife of Lir, who turned his children to swans, and was turned to an air-demon, see ‘‘ Fate of Children of Lir’’ (Atlantis, vol. iv, p. 119). Her father was reared in the god Bodb Derg’s sid-mound. 2 Aibinn, or Aibhill. See Folk-Lore, vol. xxi, pp. 186-7. Aibinn’s name, like that of ‘* Crom Dubh,” or ‘‘ the Daghda,” is concealed under an epithet, the ‘* pleasant’’ or “lovely.” 3The Dindshenchas says that Clidna, Aife, and Hdain were daughters of his Ollamh (under ‘‘Clidhna’’ and ‘‘Tuag Inbir’’). There were probably Milesian attempts to affiliate ‘Aine to the Gaulish pantheon. The Hogabil legends may be the non-Milesian element. ‘Aine’s foster-sister, Becuma, was seduced by Manannin’s son (Eriu, vol. iii, pp. 151, 163). She sought refuge in Eogabal’s sidh. 4 Rey. Celt., vol. xvi, p. 152; for the crane, see Duanaire Finn, p. 118. 5 Pp. 225, 196. St. Patrick calls it ‘a complicated bit of romance” ! 5 Metr. Dinds., vol. x, p: 115. Deda was a lake-god at Killarney (Eriu, vol. i, p- 184). His brother, Febra, also honoured at Cenn Febrat, is probably divine, like him, Aine and Aife. * Roy. Soc. Antt. Ir., vol. ii, p. 36. 8 For attempts to bowdlerize such tales see those of Clothra, Nes, Cairbre Muse, and Duben (Silva Gad., vol. ii, p. 535; Roy. Soc. Antt. Ir., vol. xl, p. 184); Clothra in Keating’s Hist., vol. ii, p. 233; Nes, Rev. Celt., vol. xvi, p. 149. ® Dindsenchas, Rey. Celt., vol. xv, p. 331. 10 Duanaire Finn, p. 197. R,IA. PROC., VOL, XXXIV, SECT, C, {9} 56 Proceedings of the Royal Irish Academy. sleep in the west of ‘Aine, daughter of Gailian, what time she fared with Dubthach,” which recalls “the wife of Dubthach,” reverenced in the cemetery- sanctuary of the Ernai, on Cenn Febrat, not far from ’Aine’s Hill. This suggests that various legends differed as to “Aine’s father,’ calling him Eogabal, Durgabal, Gailian, and Manannan. Of the other divine occupants of the sanctuary hill and mounds of Knockainey, we find several sons of Eogabal—Feri,? Ferfi, Fere, Fermait or Fer Fidail (still remembered there, like ‘Aine), Aillen, Lu, and Fainle. Another supernatural lady, Bacht (the Cacht, wife of Eogabal, of the poem), came from Sid Cliath to Drom Fingin (Kilfinan) to confer with Fingin mac Luchta, King of Munster,’ and we hear of Emer, wife of Uainide, and Eter, wife of Feri; perhaps the lady Etar, connected with Ben Edar, in other poems. In the North of Ireland ‘Aine is the Cailleach Bheara, who, at Knockainey, is reputed to have made the curious casdn, or causeway, over the Cammogue. Donn, under many epithets,‘ is reverenced in many places, as in the sand-hills of the inddir, at Lehinch, Co. Clare, and at the fairy- cairn of “Strickeen,” on the summit of Knockfirina (under the name of Donn Firinne) visible from the cairn of ‘Aine. Tue Gops’ Forts. Many Celtic gods were patrons or makers of forts.’ The god Lug gave his name to three places called Lugdunum or Lougdounon. Lyons and Leyden show a shadow of his name, and his August festival, the Irish Lugnasad, was observed at one of them.‘ Camulos gave his name to Camulodunum ; his equivalent, “Segomo,” bore the epithet “dunates” (of the dun) from earthworks.’ Nuada, Nodens, or “Ludens,” held the British fort, Dinn Llud, and probably Ludgate in London. In Ireland the difti- culty is to select analogies, not to find them. The Dagda, with his huge ! Variant parentage was common in Irish allusions to gods and heroes: cf. the war- goddess (Rey. Celt., vol. i, pp. 330 sqq.), Finn’s divergent genealogies, and Macha. Also see Rey. Celt., vol. xvi, p. 46. Nuada’s father varies in many sources. * Fere (Todd Lect. Ser., vol. xv, p. 3); Fermaise (Agall., p. 248) ; Fer Fidhail (Rev. Celt., vol. xvi, p. 152). 2 R. Ir. Acad., Irish Texts Ser., vol. i, p. 9, from Book of Fermoy. ‘ Cath Finntragha (ed. Meyer), pp. 1, 13-15); Agallamh, p. 225. For his connexion with Knockfirina see Lewis, ‘ Topog. Dict.”, p. 114; Introd., Feis tighe chonain, p. 93 ; and N. Munster Arch. Soc., vol. i, p. 108. 5 See Rhys, Hibbert Lect., vol. iv, for Lug, Nuada, and Camulos; also ‘* Irish Mythological Cycle” (De Jubainville). ©On_ August Ist, re-dedicated to the Emperor Augustus, ‘‘All Gaul met at Lugdunum.” 7 The term ‘‘ Dumiates’’ or ‘‘Dumias” (Rey. Celt., vol. xxxili, p. 463) probably refers to the Puy de Dome. Wesrropp—Ancient Sanctuaries of Knockainey und Clogher. 57 fork, cut the double trench of Lorg an dagdae. He also fortified Raith Breise, and built the Griandn of Ailech, which latter was also sacred to the war-god Neit—“ Ailech Neit, on account of Neit’s splendour.” Oengus, son of the Dagda, made a dim and dindgna, with a lofty sonnach or palisade.2, Manannan dwelt in Din Inbir; his father, Lir, in Sid Fiondachaid din;* Lug, in a rath; and the war-goddess in Lis na Morrighna, or Maiste, the great fort of Mullaghmast.t Nuada left his strong fort of Almha to his son Tadgh.° Balor, the demon-god, had his fortified promontory of Din Balor, on Torry Island,’ as the four-headed Suantowit had his entrenched temple on Arcona headland. (Borlase recalls “ Ardchonain” on Torry.) Imlech, descendant of the god Nuada Argetlamh, “built” forts in Imlech-Fir Aendarta (called after him), now Emlygrennan, at the foot of Sliabhriach.? The Tuatha Dé dug Rath coraind,*® and the horned, bull-headed god, Buar Ainech, like the Daghda, was rath-builder to Bress.* Little doubt that in the three rings of Dunainey, on the Knockainey Hill, the triad of gods, Eogabdl, Feri, and ‘Aine, were supposed to dweli; and they had also their cairns, of which ‘Aine’s one was evidently called Sid cliath, the sacred mound of Cliu. It and Uainide’s cairn are extant. Sip Mounns. The perpetual difficulty of Irish archeology is the vagueness of its nomenclature.” As the ring-forts of earth and stone and the promontory and square forts are called dun, liss, rath, dangan, port, and cathair, so the word sid connotes no type. Tumulus, or feudal mote, dise barrow, or house-ring cannot be distinguished by external features alone ;"" even excavation often 1 Battle of Moytura, p. 187 ; and Agallamh, p. 151. 2 Agallamh, p. 11. 3 Mr. H. T. Knox describes the curious Cashelmanannan at Rathcroghan (Roy. Soe. Antt. Ir., vol. xliv, p. 26). For Lir see ‘‘ Children of Tuireann” and ‘‘ Children of Lir” (Atlantis, vol. ii, pp. 115, 125; vol. iii, p. 386; vol. iv, pp. 115, 145). 4 Acallamh, p. 217; Ivische Texte, iv, 1, 4812, 4818. 5 Tadgh, see (Harleian MSS. 502, p. 148b; Book of Leinster, f. 320 d; Book of Lecan, f. 407). See also for Tadgh’s fort Harleian MSS. 5280; Atlantis, vol. ii, p. 120. ®§ Identified with the mythic High King Tigernmais in early sources. Ulster Journal of Archzeology (old series), vol. i, and O’Donovan’s note on Balor, Ann. Four MM., vol. i: cf. Borlase, Dolmens of Ireland, vol. iii, pp. 1087, 1164. * Supra, vol. xxxiii, p. 470. 8“ First Battle of Moytura,” Eriu, vol. viii, p. 31. © Book of Leinster, f. 50, col. 1. 10“ Sidh and Dun” (Echtra Nerai, Rev. Celt., vol. x, p. 221), ‘‘a dinn, a dun, an admirable (i.e. ceremonial) hill’ (Book of Ballymote, p. 490), ‘‘ Rath and Sidh Cruachan ” (Rey. Celt., vol. xvi, p. 463), ‘‘ Hunting mound, residence, outlook, and burial-place”’ of Duma Selga (idid., p. 471). 1. To assert that it can be done is very injurious to sound archeology, and encourages people unacqainted with Inish records and remains (and even some Irish writers) to indulge in baseless theories. (9*] 58 Proceedings of the Royal Irish Academy. proves that the same structure was a house and a tomb. Doubtless, the temple and tomb were often identical, and the mighty dead became a god for ages, and reverted again to be a dead man. S/d was used for haunted knolls of rock, like Crag an Aibhill, above Killaloe, “ Campul na Muckagh,” and the Sidhéan on Inishturk} Cnoe Dabilla, or Rockabill, off the coast of Leinster, and the Bull Rock, “ Bui,” off that of Munster.* The rocky hill of Almha was a std, and that of Cashel was called s#d druim. The mote-like high mound was also s/d, like the most famous one at Raith Cruachan, in Connacht; Magh Adair, Co. Clare, and Clogher, in Tyrone, are most probably congeners. The great s/d bhrug of the god Oengus was Newgrange tumulus, and those of Knowth and Dowth, near it, were dedicated to Boadan and the shepherd of Elemair, the god who once owned Newgrange. Emania had a sfd* (presumably the flat-topped mound levelled in Victoria’s reign), and Tara had one to the north of the well Neamnach and the stream Nith.‘ Ring forts, “ unroofed green raths,” could be s/d, like Sid Findachaid.* Like Knockainey, Carbury Hill, or Sid Nechtain, had a mound and two rings, evidently of the triad of Nuada, Nechtan, and his wife Boand, the Boyne.* Most probably the conjoined rings of the Clogher group were the sidh of Nechtan and his wife Cail. The diminutive term sidedn was equally vague. The “sheeaun” on Inishturk and Croach an tsidedn,in Co. Clare, were fortified rock knolls. Lissateeaun (Lios an tsidedn), not very far from the last, is a shaped natural hillock ; so is Barrinagh in Co. Roscommon.’ The sfd bhrug had sometimes a well, pillar, and faithche or green near it.* Magh Adair mound has an enclosure and one (perhaps two) pillars. Lissadeocha, near Bruree,’ has a small pillar on its summit, and Newgrange had another. As to the conjoined rings, the very complex ones seem peculiar to Co, Limerick. Cooloughtragh, on Slievereagh, Dunainey, on Knockainey, and ? Proc. R.I. Acad., vol. xxxi (Clare Island Survey, Part Il), pp. 50, 51. > Bui, see L. na hUidhre and R. Soc. Antt. Ir., vol. xi, pp. 184-5; Cnoc Dabilla in Dindshenchas. *“*Sidh bhrug Emna Macha,” perhaps a holy place of the war goddess Macha (Onomasticon Goedelicum), Rev. Celt., vol. xv, p. 463, vol. xvi, p. 46. ‘‘Emhain, the Tuatha De came to it from the west " (O’Grady’s Catalogue of Irish ss. in British Mus., Egerton iii, p. 404. * Dindshenchas, Revue Celt., vol. xvi, &c. * **Children of Lir” (Silva Gad., vol. ii, pp. 476, 552) ; ‘* Atlantis,’’ vol. iv, p. 145. * « Boyne and Blackwater,” p. 29. *For these see Proc. R. 1. Acad., vol. xxxi, Part Il, p. 51; R. Soc. Antt. Ir., vol. xxxv, p. 345; vol. xlv, p. 48; vol. xlv, p. 307. * Irische Texte, vol. iv, pp. 229, 255, 258; ‘‘ Feis tighe Chonain ” (Introd.), p. 88. ® Supra, vol. xxxiii, p. 490. Wesrroprp—Anctent Sanctuaries of Knockainey and Clogher. 59 Doonakemna, on Barna Hill, have respectively four (or five), three (or four), and three rings, three always in line.’ Killulla, Co. Clare, has two conjoined and a third joined to them by a straight earthen mound. ‘The conjoined two rings are more common. ‘hey occur at Tara, Uisnech, and several presumed ‘Oenach sites, Clogher, Monasteranenagh, and Cahermee.* A good example is found in Controversy, Co. Tipperary ; three, with a circle and a shield-shaped annexe, are near Quin, Co. Clare. As to the disc barrow, such as Cooloughtragh and Dunainey seem to be, it seems fairly certain that the demolished Zreduma Nesi at Tara consisted of three conjoined mounds, not concentric rings. Near it was an evident dise barrow, a flat ring with a little central mound (cnocan), called ‘‘The shield of Cu Chulaind.’* He (as we know) was son of “ the goddess Dechtire” (some said by the sun god Lug), and nephew of the “earth god Conchobar,” son of Ness, who was reverenced at other rings, near his “Shield”? and “ Head and Neck” mounds.° The connexion of other rings and mounds with the cultus of the di/e trees® lies outside the present subject, though there was probably a sacred hazel grove at Knockainey. MoprrRN REVERENCE OF ‘AINE. Condensing from my own notes’ and the full and valuable ones of David FitzGerald in 1879,* I must give a short account of the folk-lore and observances of Knockainey, leaving others to complete them. ‘Aine was a 1 A similar triple earthwork is shown on the old map (O.S. Carlow 16) near Bagenals- town in Kilearrig. It only appears as double on new map. 2 One mound near Cahermee fair green contained a cist ; so did Knockaun Liss, near Mallow racecourse. The fair and race probably each represented an early oenach, as do the fairs of Cush, Aine, and Knocklong. 3 Proc. R.I. Acad., vol. xxvii, p. 231, p. 379, and Plate xxxii. 4 Rey. Celt., vol. xv, p. 287. 5 See Dind. S., Rev. Celt., vol. xv, p. 288. The cairns lay north and south, like the Cush and'Aine rings. The king was forbidden to go round Tara deisiol (Bruden Da Derga, ed. Stokes, p. 19). The inauguration ceremonies at Tara were very archaic (Eriu, vol. vi, p. 134). For the ‘‘ goddess Dechtire and the earth god Conchobar,” see Book of Leinster, f. 135 b, Rhys, Hibbert Lectures, vol. iv, p. 143. Cuchulaind was reverenced at Tara and perhaps at Sid Setanta in Muirthemne (Ivische Texte, i, p. 215). 6 For the bile trees, see Ann. Four MM., 982, 1051, 1099, 1111, 1143. The Bile Buada (Atlantis, vol. ii, p. 102). Bile ratha (‘‘ King and Hermit,” Meyer, 456-7). Also see Imram Bran, ‘‘ The Voyage of Bran,” p. 57; Ancient Laws, vol. iv, p.143; Tain bo Flidhais (Celtic Review, vol. iv, p. 23), and Dindsenchas (Rey. Celt., vol. xvi, p. 277). * T heard of ‘Aine and the red bull, the meadowsweet, and the Garrett Earla legends at my old home, Attyflin, farther north, about 1870, from the peasantry. 8 Rey. Celt., vol. iv, pp. 185-191. See alsw vol. xiii, p. 435. Nicholas O’Kearney in R. Soc. Antt. Iy., vol. ii, p. 32, and Introd. to ‘‘ Feis tighe Chonain,” pp. 93 and 169, says Miluachra, ‘Aine’s sister, is ‘‘ The Cailleach Bheara.” His theories are unreliable in some cases. , 60 Proceedings of the Royal Irish Academy. water spirit, and has been seen, half raised out of the water, combing her hair. She was a beautiful and gracious spirit, “the best-natured of women,”! and is crowned with the meadowsweet (spira), to which she gave its sweet smell. She is a powerful tutelary spirit, and (aided by Cleenagh and Aibhell) attacked and repelled the grim Atropus from a sick man.? She was connected with the moon, for her hill, “Carron Hill” (recte cairn), was “sickle”-shaped and meant “ sickle hill,” and men, before performing the ceremonies, used to look for the moon (whether it had risen or not), lest they should be unable to return, They used to go in procession on St. John’s Eve, carrying lighted wisps (cliar, whence “’Aine Cliar,” recte cliach), encircling the hill, and visiting “ the little moat called Midlach eruachain laimhre leab’ an triuir,” the mound of three persons* (perhaps the local triad), visiting the tillage and — meadows to bring luck to the crops and cattle.* Some girls, who lingered behind one time, saw ‘Aine, who asked them to go, and showed them, through a ring, that the hill was crowded with fairies. Once the procession was not made, out of respect to a dead neighbour, but the fairies were seen marching round the hill-top. ‘Aine was violated by the Earl of Desmond (a recollection of the legend of Oilioll Aulom). She had a son, Geroid Iarla, “ the magician Earl of Desmond,” who is still seen riding over (or glittering under) the ripples of Loch Gur till his horse’s golden shoes are worn out. She owned a red bull, with which she entered the “ green hill.’* “She, or the Cailleach Bheura,” laid the great blocks of the casdn over the river. Indeed the Sliocht ‘Aine, several local families,’ descend from her in the female line. As to the procession round a mound or cairn, the King of Tara was forbidden to go deisiol round that sanctuary, though other persons performed the rite there between two cairns. The Agallamh has “around the brugh let him walk deisiol.” A poem, A.D. 555, ' This trait appears in old literature, as where she befriends Becuma (‘‘ Adventures of Art,” Book of Fermoy, p. 139, Eriu, vol. iii, p. 163). ? See poem of 1737 in Introd. ‘* Feis tighe Chondin,” pp. 93-169. Cf. Cuchulaind’s fight with Hercules, and the defeat of the Tuatha Dé by the Philistines. In later days absurd pseudo-classic ‘‘ ornament” is common. *So Tara is ‘‘ Tulach an Trir,” from the ‘‘three fairy youths” of Eman (Battle of Magh Leana, p. 92). It was forbidden to disturb a meeting-tulach, or quarrel in a fort (Anc. Laws, vol. i, p. 175). ‘Compare the cultus of the Matres in Gaul, deities “of growth and fertility,” ‘* worshipped at large central gatherings during the great festivals ” (MacCulloch, loc. cit., p- 279). ° Rey. Celt., vol. iv, p. 191. Or, as more usually said, ‘‘ when the silver shoes of his horse ” are worn out. ® She was called the ‘‘ Lady Awney,” of “Orange,” in 1870. In some legends ‘‘Orange ” is Norway. I have heard of ‘‘ King Liamgowl of Orange ” landing in Waterford to fight King Shamus, and the Danes came from Orange. 7 O’Briens, FitzGeralds, Dillanes, Creeds, Laffans, and O’Deas. Westroprp—Ancient Sanctuaries of Knockainey and Clogher. 61 mentions “ the hosts that go round the cairns”; and long before the days of Medb, Poseidonius tells how the Gauls made a turn to the right. The reverence for the monuments is well shown by the tabu in the Book of Rights not “ to sit on the tomb of the wife of Maine”; and another, in the Agallamh, “not to walk on the sidh of Femen (in Co, Tipperary, the residence of the god Bodb Derg), by the new-kindled blaze of a red fire,” like the eliar at ‘Aine. In the same work we are told that men feared to sit on the three tulachs on Cean Febrat (till St. Patrick did it), for fear of the Tuatha Dé Danann. I may add the oft-cited case at Croaghateeaun, where we were told “to cross ourselves against the Dannans.”! It should be remembered that the god of the Galtees, the harper Cliu, came himself from Baine’s sidh mound. THE REMAINS. The long, low hill (as the “ Mesca Ulad” notes) has a noble outlook in all directions. The view from such a place should be studied, and it is note- worthy how many prominent points in legend and fairy lore are in sight— the dolmen-topped Cromwell Hill and Duntrileague, the cairn-capped Knockfirina, and the Loch Gur hills, the castled ridges of Knocklong and Rathcannon, the fairy hill of Sid Asail, the flank of the cemetery of Clann Deda on the edge of Cenn Febrat, and the boundary ranges named by the Mesca Ulad. It is boldly curved, and the marshy fields and pools at its south- west foot mark an ancient lake,° full when the woods of Coill Moér and Coill Cruaidh condensed the mists of the mountain tops, and poured them into the now shrunken Lubach, Saimer, and Cammoge. It was also once fenced by deep marsh to the east. The Grossi Fines, Roll No, 14, names the Lake of Any as existing in 1522. THe CatrNs.—That of ‘Aine, the Sid Cliach, occupies (as Cormac’s Glossary’ says), the highest ground, standing on the eastern brow 537 feet above the sea. Though so famous in legend and lore, it is a defaced, insignificant heap of earth and stones wrecked by treasure-seekers,* daring, I 1 Rev. Celt., vol. x, p. 287; Ann. Four MM., 555 ; Book of Rights, p. 5; Agallamh, p- 125. Poseidonius, quoted by Athenaeus, Lib. iv, p. 152. Forts were used for magic. See ‘‘ Rath of the Incantations””’ and the ‘‘ Duma nandruad ” in Irische Texte, vol. ivy, 1, and Eriu, vol. viii, p. 49. 2 Dindsenchas, Rev. Celt., vol. xv, p. 441, Crotta Cliach. He died of fright when Baine, in the form of a dragon, burst out of the mound at his magic harping. 3 Like Loch Lungae (Trip. Life, p. 209), Loch bo (Agall, p. 123), Loch Cenn, and other lake sites. Much of Loch Gur and Monaincha Lake and Coolasluasta Lake have been drained away in modern times. +Sanas Chormaic, p. 9. 5 No new outrage ; not only the Norse but the Irish ravaged such monuments. See many interesting accounts in Agallamh. 62 Proceedings of the Royal Irish Academy. presume, from a reliance on ‘Aine’s proverbial kindness. Much, too, was taken for the long wall beside it. It measures 48 to 55 feet across, and is 11 feet high to the west, 6 feet to the south, and 8 to 9 feet elsewhere. No large stones remain. It is beside an apparently artificial hollow, dug along one of the great rock-ribs so characteristic of the hill. A faintly marked enclosure and an even fainter circular hollow adjoin the cairn. Tue Ecuiasa PILLArs.—No trace remains; they may have been used for material for the castles and other buildings in the village below the hill. THE ConJornED Rincs.—'This is called Dunainey and the “ Mullach an triuir” ; three fosses, with rounded mounds of the Rathnarrow type, one in each ring. A modern fence crosses them, like that through the Coolough- tragh rings at Cush. They were probably disc barrows, and sepulchral ; perhaps, like several of those excavated, they were women’s graves.2 But for its poor preservation, the whole would be comparable with the Cush monument, No trace remains of an enclosing mound round the whole; if it existed, it was possibly 6 feet wide, as the fosse rings are 12 feet apart. They lie north and south; the northern a little to the east of the central axis of the others. The whole measures (if we include the northern cairn and ring 27 feet away) about 250 feet north and south. The fosses and interspaces are 12 feet wide. The rings are 33 feet, 36 feet, and 36 feet, taking them from north to south, and about 63 feet to 54 feet over all, east and west. The cairn to the north of them lies in a ring 6 feet thick, and 63 feet over all; it isa heap of small stones 18 feet through and 4 feet high, the centre opened? The rings were deliberately dug on a slope, the edge of the southern touching the edge of the plateau on which ‘Aine’s cairn stands, 99 feet away. O’Donoyan did not notice the northern rings; he took no interest in the the fort, and ” earthworks, and contemptuously notes the “ defaced cairn, “two small mounds evidently sepulchral,’ apparently the southern rings. He adds that “’Aine was still (1840) supposed to haunt the hill in the shape of a banshee.”* ' For ‘‘a district marked by a stone of worship” see Ancient Laws, vol. iv, p. 143 ; the largest circle of Lough Gur has been removed since 1850. 2“ Ancient British Barrows” (Thurnham in Archaeologia, vol. xliii, pp. 285, 348) ; “Ancient Hist. of South Wiltshire” (Hoare, 1812), p. 21; ‘* Tumuli Wiltunenses ” (same, 1821), pp. 19, 159. 3 Keating cites from Book of Lecan, f. 258, in ‘‘ Three Bitter Shafts of Death,” the passage, “a small rath was raised round the corpse, with a leacht, or cairn.” This accurately describes this monument. 40.58. Letters, vol. i, p. 229. Wesrropp—Aneient Sanctuaries of Knockainey and Clogher. 63 Tur Rina Forrs.—The western, perhaps the dwn of Conall Eachluath, repaired by King Brian 1002-1012, lies on the western brow, above the wooded demesne of Kilballyowen and the O’Grady’s resideuce. It may be the earthwork attributed to Eogabal “on the west.’ It is crossed by a fence, and much injured by cattle, being of red, friable earth. ‘here is some trace of a fosse, especially to the north, where it is 12 feet wide below, and has a spring in a rock-cut tank filled with water-plants. It is high and steep to the north, and is 125 feet across, north and south, and 108 feet east and west. The ring is 12 feet thick, 10 feet high to the south, and 12 feet to the north, its base 12 to 18 feet thick. The platform is 5 feet above the field, and the whole about 177 feet over the fosse. The southern fort seems to have been largely of stone; slight trace remains beside a cottage on the south slope, near the ancient track, up the middle of the ridge, where it joins the by-road called Bohernaskagh, “ the way of the thorn bushes,” leading westward from the ancient fair green. There is also a curious well half way down the slope, south from ‘Aine’s Cairn, cut in the rock, but with only a small flow, even in wet weather. The rectangular cleavage of the rocks makes it very regular; it is partly fenced with large rocks torn out of the crag. The casén or causeway is of large slabs, bridged over by ones of considerable size. Mr. H. S. Crawford has illustrated and described it for the Journal of the Royal Society of Antiquaries of Ireland, vol. xlvii, p, $2. ‘OENACH CULI, CLoGHER (O.S. Map, No. 42). When I first noticed the unusual group of remains in the townlands of Clogherbeg and Raheenamaddra, I saw that the place had been of ancient note, and the name suggested ‘Oenach Clochair. This, however, had been located as near Monasteranenagh, and the number of ‘“ dogmatic equations” published about Co. Limerick throw unusual difficulties in the way of topographers. The whole question had to be reconsidered, and it was only after I had got the evidence methodized that I ventured to assert a new identification. 1 overlooked till Jater the fact that Mr. P. J. Lynch had already suggested the same ;' but as he gives no reason for it, I will here give the evidence, at some length, which led me to the same conclusion. ‘Oenach Clochair, or ‘Oenach Cili, was one of the chief assembly places of the Munster men, and the cemetery of the Dergthene,* as Cenn Febrat appears to have been of the Ernai. Our predecessors followed O’ Donovan in regarding 1 Journal North Munster Arch. Soe., vol. ii, p. 10. 2 «Tract on the Cemeteries” given by Petrie in ‘* Round Towers.”’ R.A. PROC., VOL, XXXIV, SECT. C. {10} 64 Proceedings of the Royal Irish Academy. it as ‘Oenachbeg, somewhere near Monasteranenagh Abbey. ‘Two early documents, however, should have taught them better. The Mesca Ulad' represents the invaders coming southward, towards Cenn Febrat, from Knockainy, to “’Oenach sen Clochair,’’*? where they camped, though “the rough winter weather was not the time of the ‘Oenach.” The Agallamh,’ so rich and accurate in its topography, tells how St. Patrick left Cuillend ua Cuanach (Cullen, in Co. Tipperary, on the border of Co. Limerick) and, on his way to Ardpatrick, skirted ‘‘the denach of Nechtan’s wife, Cuil, now called the heifer-carrying ‘fair’ green of old Clochar (6enach sen Clochair).” Now Monasteranenagh lies far to the north-east of ‘Aine, quite off the route of the raiders, and, similarly, the saint was not likely to have been supposed to make a huge circuit, for no recorded object, when he wished to reach Ardpatrick. Take a line from Cullen to Ardpatrick on the map, and it touches Clogher. There can be little doubt from the Irish sources that the latter, and not “‘Oinach orbece” or “’Oenach beg,” is intended. “ Finn and the Phantoms” does nothing to fix ‘Oenach Clochair, but is valuable as showing that horse-races took place there. “Munster men from the plain gathered at it; they ran three clear races at the faithche mic Maareda.” The black horse of Dil, son of Dachrech, ran “to the rock over Loch Gur,” and King Fiacha Muillethan asked his maternal grandfather for the steed. The Norman records seem equally clear; Prince John’s charter‘ to De Magio Abbey, or Monasteranenagh, in 1186, confirms to it “ Enachceuli in Corbally,” and mentions Clochur and Loc Geir (Gur). Corbally townland adjoins Raheenamaddra, in which the remains occur, and is less than a mile distant from the mounds. Corbali, in Grene, Kilfrush, and Dunmoon, all in the Clogherbeg district, were held by the Hospitallers.’ Enachculy is also named in a lawsuit at Limerick of Maurice de Londres* against Monastera- nenagh in 1254. The god Nechtan, it will be remembered, was reverenced with his wife Boand and “the great god Nuada,” at the source of the Boyne.’ He also sent Eogabal and his family to Knockainey; so Cuil, his other wife, was 1 Ed. Hennessy, p. 19. * The ** sen” shows that even when the Mesca Ulad was written the antiquity of the place was realized. 3P. 118. * Rey. Celt, vol. vii, p. 29, Duanaire Finn (ed. Ir. Texts Soc.), p.127. There was also a ‘‘ faithche mhic Maireda "’ on Magh Eala. * Cal. D. Ireland, No. 136 ; see also 96, 2138, and p. 288. ° From these ‘‘ Londoners” Ballylanders, near Clogher, is named. * Sir W. Wilde's ‘‘ Boyne and Blackwater” (2nd ed.), p. 29; Metr. Dinds., vol. x, pp- 27, 29, 31. Westrropp Ancient Sanctuaries of Knoekainey and Clogher. 65 probably a local goddess, reverenced with ceremonies and races at ‘Oenach Culi, or-’Oenach Clochair, by the Dergthene, who reverenced Nuada, Of course ‘ fair” is a most inadequate translation of the ancient ‘ “Oenach.” The remains suggest a place of ancient worship, and resemble other pagan cemeteries. At “ Brugh” we have three great tumuli, several small ones (like the Clogher mounds), ring works, and pillars, also wells and streams. Mounds were there, one called “the Dagda’s Head” (like “ Cuchulaind’s Head ” at Tara); others as “‘ the Comb” and ‘ the Casket of the Dagda’s wife,” and the Duma Tresc ; rings like the Fert of the Dagda’s son Aedh; twin mounds like the “ Two breasts of the Morrigu.” So at Tara we find great conjoined works like the Teach Cormaic and the Forraidh, large disc barrows like the “Shield of Cuchulaind,” conjoined mounds like 'l'reduma Nesi, mounds like the Duma na ngiall, and lesser ones like Dal and Dorcha; ring-forts and pillars, wells and streams, and tracks. ‘Oenach Carmain had seven mounds where the dead were bewailed, twenty-one raths, a cemetery, and three markets. A large track leads southward from Ushnach, which has its cemetery, conjoined rings, and. its holy rock and springs. Cush, the Cenn Febrat cemetery, has ring-forts, conjoimed rings, large and small tumuli, tracks, springs and streams, and pillar-stones. The Clogher group is closely similar. THe Remars. The site has three (or four) rising grounds, hardly hills; the northern called Clogherbeg, the north-eastern Clogher Hill, the south-western Knock- aunatarriff, or “bull’s hillock,”! the south-eastern nameless. Clogherbeg has on its summit a small perfect mound, a bowl barrow, with no fosse, 7 to 11 feet high, and about 99 feet across the base. Near the road to Knocklong village is Raheenamaddra, the dog’s little fort, a flat-topped mote, or barrow, 10 to 14 feet high. An old raised causeway, farther on, leading from the north-east to Knocklong Hill, is also noteworthy. Turning southward, past the “Cross of the ‘'ree”* and over a brook, we enter the fields to the west. Crossing a very regular depression, probably natural, and very likely the limit of the denach eastward, we find, on the eastern rise, traces of enclosures. One is a platform, certainly ancient; it is oblong, 93 feet east and west, by 61 feet; raised 2 to 4 feet over the field, 1 [t may be a casual name, but bull feasts were a feature of lrish pagan assemblies. For the name Donnotaurus and its connexion with the ‘‘ Donn Bull” and *‘*'Tarvus trigaranus,”’ see MacCulloch, Dict. Relig. and Ethics, vol. ili, p. 296, ef. p. 294; and men masquerading in bull hides, ibid., vol. y, p. 838. * 'The ‘Tree ” was enclosed by Mr. Ryan, of Scarteen, 1862. 66 Proceedings of the Royal Irish Academy. with a slight fosse and rounded corners to the east, and a hollow track along its south side. From the latter side two parallel tracks, also 6 to 8 feet wide, run towards the south-east to the hollows or dry pools 84 feet away. Between these ponds is a mound 4 feet high, of doubtful age. From the double track another one leads at right angles towards the west. Crossing the stream and valley, and going up Knockaunatarriff, through a marsh thick with yellow iris, or “ flagger,” we get a noble view of Sliabh riach and the Galtees ; the rise has a slight terrace, probably natural. ‘win Mounps.—At the end of a long “screen” (plantation), and on the summit, we find conjoined twin tumuli; they are cut by the mearing of the townland, and that of Mitchellstown Down, and recall the “ Two Breasts” at the Brugh. Each has been opened; the cross fence has been made between them, and the boundary fence cuts their west flanks. ‘They are on a steep slope, so the southern is 13 feet high, in three slightly marked stages, to the east, and 5 feet to the west, and the northern 10 to 12 feet, and 5 feet high. ‘The first is 13 feet across the top, and 45 feet below; and 15 feet from it, beyond the fence, is the northern mound. It is 15 feet on top, and over 30 feet below. Rixc-Mounp.—In the field, lower, but still on the summit, is a ring- platform, 60 to 61 feet across, and 5 feet high; a fosse 12 feet wide and barely 2 feet deep remains to the south, west, and north; there is a trace of the outer ring to the south-east. ConJorneD Rincs.—Going northwards down a gentle slope we reach a flagger patch and an old track about 100 yards from the upper ring, and at 33 yards farther the most remarkable of the remains. This consists of two platforms conjoined? within one fosse 12 feet wide. Each is D-shaped in plan, and 4 to 5 feet high, and the western is 60 feet across east and west, and 54 feet north and south. The dividing fosse is straight, 9 feet wide below, and 24 feet from platform to platform. ‘The eastern section is 81 feet across, north and south, 78 feet east and west. The main fosse is obliterated to the north-east of it. ‘The western ring has signs of large stone facing; one block is over 3 feet long and high, and 2 feet 6 inches thick. From its position any ceremony on the platforms must have been visible to many ' Ballenvistallane-Down, in 1587. Fiant 5179. ?T have noted such remains in the following counties: there are possibly many others :—Clare 3, Kerry 1, Limerick 13, Tipperary 14, Cork 2, Waterford 1, Louth 1, Meath 5, Westmeath 1, Carlow 1, Roscommon 4, Mayo 1. The preponderance in Limerick and Tipperary is very marked. “MOIMANI] (OD SXNHDOTD GNV AANIVHOONY AO SATYVAIONVS INAIONY—‘ddOULsa AY ° 13344°31W9S | a x On BB ‘tqdad pur ouly vqebo ‘S jy) UNAS 40 ot sdeyaad “ “sSury paurofuos NY le S Mt! Wy y SS aoyy Gaqaayso ww Il TIS lly “nny ou W un Q “ayy » ANS 7, UMO UMOAS(|>423 1 ail! ) WW %G D umn MN wes QW Y MW gz Ven, B Se , ES —— 4 oot O01 ° 2433440 31V90S apruiey| jo, Usted YqION Z uivazeu yu ZZ v alae uiw320]9 2vuaQ *SONIY GINIOFNOD dNOYD Y3HOO19 SY" faqsay5079 “9 “LOUS ‘AIXXX “TOA “avoy “I “YL ‘90%d *] FLVIg Westropp—Ancient Sanctuaries of Knockainey and Clogher. 87 hundreds of spectators on the slope. South of the main fosse is another, perhaps part of an old hollow way leading to it; the section leads down to a hollow or pond not far to the south-west of the mounds, and has trace of an outer ring between the fosses 9 feet thick and 2 feet high. Eastern Rinc.—There is an oval ring platform to the south-east about 150 feet away. It measures 66 feet north and south to 81 feet east and west, rarely 4 feet high; the fosse is 10 feet wide and 2 feet 6 inches deep. An old track, 45 feet from it towards the conjoined rings, runs between the “forts.” To the south-west, 14 yards away, is a shallow, oval basin, 18 to 23 feet across, and about 4 feet deep. Beyond it, 120 yards from the fort, is a large stone over 4 feet long, 3 feet high, and 8 feet 6 inches thick. Another, possibly as large, but partly buried, lies to the north-east above the slope to the stream, which joins a second stream, and has several pools and springs beside it. ‘The tracks may be ditches of old fences, but have no traces of mounds, and in most eases lead direct to the earthworks. They are fairly uniform hollows, rarely over a foot deep, and 6 to 8 feet wide. They are, I think, old tracks, like those on Slievemore, Achill, and those leading to the Black Castle causeway at Loch Gur and the promontory forts of Doonaunroe, Dundoillroe, and George’s Head, Co. Clare, and Clashmelchon and Pierce’s Island, Co. Kerry. I saw no other remains in the ancient chief cemetery of the Dergthene, but many must have been destroyed by cultivation in the tilled lands of Mitchellstown Down, on the same ridge. Between farming and fox coverts it is wonderful that any remains escaped to our days. I trust that this paper, despite its limitations, may prove of use to those studying the obscure and neglected subject of our early sanctuaries, and lead others to devote themselves to the identification of other sites of pagan worship in our island. 1 The Ancient Laws of Ireland, vol. v, p. 475, ‘‘ erecting stands and platforms at an ‘Oenach.” Vol. iv, p. 220, ‘‘ the cladh of a fair green.” Vol. i, p. 233, disturbing a fair green ; p. 129, cleansing it for sports. R.I.A. PROC., VOL. XXXIV, SECT. C, | 11] [ 68 ] ve ON SOME INTERMENTS AT MOORETOWN, CO. MEATH. By R. A. S. MACALISTER, Litt. D., anp J. R. D. HOLTBY, MB. PuateE II. Read Fesevary 12. Published Aveust 27, 1917. Mr. JosepH DoLan, Ardee, wrote to me three or four months ago to the effect that a find of human bones had been made in the course of quarrying operations at a site near his town. Mr. Holtby and I took an early opportunity of visiting the place under his guidance. The site is in the middle of a field on the townland of Mooretown, just over the boundary of Co. Meath. Mr. Moore, the proprietor, made us welcome, and gave us every help in his power. We found that in digging down to the rock for quarrying purposes the workmen had cut through a series of shallow graves, each containing at least one skeleton; but the bones had much decayed, and were in a very rotten state, and a good deal of damage had been done by treasure-seekers in the interval between their discovery and our arrival. The osteological material recovered was therefore not so much as it might have been. The graves were shallow trenches, about 2 feet 6 inches below the ground level, lined on each side for their lowest 10 or 12 inches with slabs on edge, supporting horizontal cover-slabs. The length of a perfect grave exposed was 8 feet 3 inches, and its breadth 1 foot 6 inches. There were eleven graves exposed when we came to the ground. Most of them had been broken and were empty. They lay approximately east and west, the heads to the west; but they showed a tendency to radiate ; the last four of the series pointed more towards the north. Nothing whatever was deposited with the bodies to give a clue to their age. Of the eleven graves exposed, the first three (counting from the south) were empty, the fourth contained an astragalus, the fifth and sixth were empty, the seventh contained the remains of a man’s skeleton, the eighth the fragment of a skull, the ninth part of the bones of two women, the tenth and eleventh were empty. A twelfth grave, we were informed, had existed south of No. 11, but had been destroyed by the quarrymen. We tried with a Macaisrer AND Hoirsy—ZInierments at Mooretown, Co. Meath. 69 crowbar and at last found another grave, west of No. 4; it had, however, been imperfectly lined—there were stones round the lower part of the tibiae only—and the rest of the skeleton, unprotected from the rich, damp soil, had absolutely decayed to nothing. R. A. S. M. As regards the human remains, portions of two skulls, a femur, and two tibiae were recovered, together with a talus from a young subject. ‘I'he skulls, which we will distinguish as A and B, were unfortunately so damaged as to render impracticable the estimation of many measurements which would have thrown a fuller light on their characteristics. Both skulls were from male subjects, and were dolichocephalic, though not to a marked degree. The chief measurements are appended. It will be noted that the index of cephalic length is 71 and 70 respectively; figures below 75 aie considered to indicate a long-headed condition. The altitudinal index in these cases was estimated by a method which I have recently described, actual measurement of the height being impossible owing to absence of the cranial base. In skull A the difference between the cephalic and altitudinal indices was rather more marked than is usual. It is, however, possible that, owing to the imperfect condition of the specimen, the cephalic index should be a little higher than the figure given. The left occipital fossa of this skull showed a very well-marked elevation, indicating a definitely operculated and defined sulcus lunatus (Affenspalte)—a condition generally seen in, though not confined to, primitive races. The teeth showed the well-worn condition of the crowns so common in ancient specimens. The petrosquamous suture was more obvious than is usual. MEASUREMENTS AND INDICES OF SKULLS. SKULL A :— Maximum length (glabella to occipital point), 18 em. Maximum breadth, y - s ; ee L2:crem: Cephalic index, 3 : ; : ; Prva (ical Opisthiobregmatic height, . ‘ ‘ ee Lopecmn. Altitudinal index, ‘ : : ' 6 Skutt B:— Maximum length, 1 Maximum breadth, : : : od? Cephalic index, .. ; : : en 20 Opisthiobregmatic height, 1 Altitudinal index, 7 70 _ . Proceedings of the Royal Irish Academy. - Live Bones. Femur.—This showed distinct anteroposterior flattening at its ve end, as indicated by a platymeric index of 67-6. The linea aspera was raised and prominent, giving a pilastric index of 107:1. The upper and lower extremities of the bone were missing. Tibia.— Portions of two were recovered. Each showed flattening from side to side or platycnemia (the so-called razor-like tibia), the indices being 63-1 and 65-7 respectively. Each also showed a distinct facet at the fibular margin of the front of the lower extremity. Talus.—One was recovered. It had belonged to a young subject. It showed a marked extension of the articular surface on to the medial side of the neck of the bone, This, like the facet described on the lower end of the tibia, is due to pressure between tibia and talus. Norr on Lirup BoNgEs. The normal index of platymeria in modern British femora is 818 (Hepburn). This anteroposterior flattening of the upper portion of the femoral shaft is commonly associated with side to side compression of the tibia, and is due to unequal repressive development of the various components of the quadriceps extensor muscle of the thigh, resulting from functional activity, such as might be necessary in the habitual assumption of the posture of squatting. It is often found also in conjunction with a markedly raised condition of the linea aspera on the back of the femoral shaft. The amount of pilastering here found (107-1) might, however, be found to-day, though insufficient figures are as yet available as to its incidence in modern hones. The pilastric index has, however, some significance on account of the other associated conditions. Platycnemia is quite uncommon in modern tibiae, the breadth index of these ranging between 80 and 100 [Manouvrier gives 88-2 as the average for modern French specimens]. It is usually ascribed to influence of strongly developed calf muscles, and is found frequently in races of squatting habits. It is very marked in these bones. Macauister AND Hoursy—ZJnferments at Mooretown, Co. Meath. 71 SUMMARY. Modern skulls tend in the aggregate, so far as this country is concerned, to be mesaticephalic, with an index between 75 and 80; and it would be quite unusual in taking a series to find any two, chosen at random, with indices such as the two in this series possess. The specimens from Cork Hill,-which could be dated as about 900-1000 A.D., had indices well over 75, and I think that we are justified in assuming that these were probably earlier. ‘I'he oceurrence of the impression for the Affenspalte strengthens this view. It should, however, be noted that as Ireland was subject to invasions by various races, such a migration might have a pronounced effect on the skulls of a particular locality. I do not know if there is such a factor to be considered in the case of the site from which these remains were recovered. The evidence afforded by the limb bones is in favour of an early date; and, so far as the remains are concerned, Dr. Macalister’s view that the burial-place was pre- or early Christian seems reasonable. J. R. D. H. R.I.A, PROC., VOL, XXXIV, SECT, C. [12] See ah - vy Wag ? a aa r . a 7 iia = . 7 | ba : . a 4 j é ‘8 eb a “ . a ry f 9 1p hy . eo ab ae fideo, AD Tobi : © pa he mained 4 : a JAE a. a Proc. R. I. ACApD., VoL. XXXIV, Sect. C. Prate IT. FIG, 2. MACALISTER AND HOLTBY.—INTERMENTS AT MOORETOWN, Co. MEATH. [Sakonly Mi THE DISCREPANCIES BETWEEN THE DATES OF GIFT AND HALL-MARK DATES FOUND ON PIECES OF PLATE BELONGING TO TRINITY COLLEGE, DUBLIN. By M. 8. DUDLEY WESTROPP. (Read Novemper 30, 1917. Published January 2, 1918.] Art the request of the authorities of Trinity College, Dublin, I made, in June, 1916, a catalogue of the plate belonging to the College. As I proceeded with the work, I was surprised to find that the date of gift inscribed on a large number of pieces was much earlier than that indicated by the hall- marks. This, at the time, I tried to account for in two ways. First.—In former days plate received comparatively rough usage and in time appeared unsightly, so that it became a common practice to have a disfigured piece remade, either in the form of the original, or in that of the period in which it was remade. Second.—The donors may have given money at a certain date for a piece of plate which may not have been purchased till a later period. On April 23rd, 1917, Dr. J. P. Mahaffy, Provost of Trinity College, Dublin, read a paper on the subject before the Academy, in which he rejected these two theories as being unsatisfactory, and also he tried to prove that the inscribed dates on the plate were the ones to be relied on for date of manu- facture, and that the hall-marks were of secondary importance. In his paper the Provost quotes from a register of Oriel College, Oxford, that it was the frequent practice to sell or exchange pieces of plate worn out or no longer in fashion, but that the inscriptions were in most cases carefully transferred. ‘I'he Provost also admits that numerous pieces of the College plate, such as cups and tankards, were in the eighteenth century melted down, and made into more useful articles, but that the names of the original donors were inscribed on the new pieces. An entry of the year 1766, in the Bursar’s books of Trinity College, Dublin, records a payment to a Dublin silversmith for the exchange of old plate. ‘hese records appear to substantiate in some way my first theory. I will now endeavour to show that the date indicated by the hall-marks Westropp—Dates on Plate belonging to Trinity College. 73 is the one to be relied on for date of manufacture. From long experience I have found that inscribed dates on plate are most untrustworthy for ascer- taining the date of manufacture. I have seen comparatively few pieces on which the inscribed date of gift and the hall-marks synchronize. This usually occurs only on exceptional pieces, such as maces or such other objects as would have been made specially for the occasion. In the case of cups, salvers, &c., which formed the usual stock of a silversmith, any piece could have been selected and inscribed with a date according to the wishes of the purchaser, irrespective of hall-marks. In Iveland the hall-marking proper of plate begins in the year 1638, although a mark for standard silver is mentioned in 1605. In 1637 the Goldsmiths’ Guild of Dublin received a Royal charter, which empowered them, among other things, to strike a punch of a crowned harp on all standard plate. Hall-marking was begun early in 1638, and at the same time an alphabetical letter, generally changed annually, was adopted and struck on plate. The alphabetical letter was struck to denote the year in which a piece of plate was made, so that in case of any irregularity the master and wardens of that year could be held responsible. ‘‘he harp crowned, date letter, and maker’s mark were the only marks struck on Dublin plate from 1638 to 1730. In the year 1729 an Act of the Irish Parliament (8 Geo. II, ec. 3) laid down that from and after the 25th day of March, 1730, a duty of 6d. per oz. was to be imposed upon all gold or silver plate wrought or manufactured which shall be imported and brought into the kingdom of Ireland, and that upon all gold and silver plate which shall be made or wrought in Ireland at any time or times from and after the 25th day of March, 1730, during the term of twenty-one years, a duty of 6d. for every ounce 'l’roy was to be paid by the makers or workers thereof respectively. Also that no goldsmith or silver- smith or other person working or trading in wrought or manufactured gold or silver shall sell, expose to sale, barter, or exchange any gold or silver vessel, unless it be silver wire, or such things not exceeding 4 dwt., until such time as such plate, vessel, or manufacture of gold or silver shall be assayed, touched, and marked, upon pain of forfeiting the value thereof ; and that after the same shall be touched and marked, the duty of 6d. per oz., imposed by the Act, shall be paid by the person bringing such plate to be assayed and touched, to the assay-master, who is hereby empowered and required upon receipt of said duty to stamp or mark without fee or reward the said plate with such stamp or mark as the Commissioners of His Majesty’s Revenue for the time being, or any three or more of them, shall from time to time appoint. ‘l’o denote the [12] 74 Proceedings of the Royal Irish Academy. payment of the duty imposed by this Act the figure of Hibernia in an oval’ punch was struck on Dublin plate. An entry, dated April 2ist, 1730, in an assay book belonging to the Dublin Guild of Goldsmiths, states that “ duty came on this day.” From this Act it is apparent that no duty stamp was struck on imported plate, but only on plate made in Ireland. Although the mark of Hibernia was struck on Dublin plate from 1730, being first found in conjunction with the date-letter L of that year, it was not legally authorized until the year 1776. In the latter year one * Michael Keating, a Dublin silversmith, was convicted of counterfeiting the duty mark on plate; but on looking into the matter, it was found that the Commissioners of the Revenue had never appointed any mark to be used by the assay-master. Michael Keating was accordingly acquitted on this charge. The Commissioners of Revenue afterwards received a full certificate of Mr. Thomas Nuttall being appointed assay-master, and they made the proper order therein, and appointed a particular mark to be used by him in the future, and to be struck on all sterling plate pursuant to the Act of Parliament. The Hibernia mark was probably selected by the Dublin Guild of Gold- smiths, though no reference to the matter occurs in their proceedings, nor is there any difference in the detail of the mark from the year 1776. Acts of Parliament were passed from time to time reimposing the duty on plate, which was finally abolished in 1890. The requirements of the Act of 1729 as to the marking of plate do not appear to have been always fulfilled, as frequently pieces of Irish plate are found, made both before and after 1730, which bear no hall-marks, but only the maker’s mark. To remedy this another Act of Parliament was passed in 1751 (25 Geo. II, ec. 10), which laid down that, “ Whereas silver plate is often sold without being assayed, touched, and marked, to avoid the payment of the duties, for remedy whereof it is enacted that from and after the first day of May, 1752, no person shall buy, take, and receive in the way of purchase, barter, or exchange any wrought or manufactured gold or silver plate of or from any goldsmith or silversmith or any person working or trading in gold or silver, unless it be silver wire or such things, not exceeding 4 dwt. not being assayed, touched, and marked by the assay-master at the time such gold or silver shall be delivered to the buyer, upon pain of forfeiting the value thereof.” From this latter Act the Provost draws the conclusion that all unmarked plate made preyious to the first of May, 1752, was to be called in, assayed, Wersrropp—Dales on Plate belonging to Trinity College. (6 and marked. The Act, however, does not state this, and makes no mention about the clause being retrospective. There are pieces of plate belonging to Trinity College, made in the second half of the seventeenth century, and not hall-marked yet. The Act of 1729 sets forth clearly that all plate made on and after the 25th March, 1750, was to be liable to the duty of 6d. per oz. - Plate made before this date was not liable to duty, and, therefore, did not bear the figure of Hibernia. If a piece of plate made, say, in the year 1720 was found not hall-marked, and was brought to the hall to be assayed and stamped, say, in 1752, it would not be liable to duty, not having been made after 25th March, 1730, and, there- fore, could not have the duty-mark of Hibernia struck on it. The Provost in his paper suggests that pieces of plate sold unmarked were years afterwards assayed and stamped with the hall-marks of the period in which they were assayed. This conclusion he draws from the fact that several pieces of the College plate have the assay scraping mark still visible. All plate to be assayed must have a small portion scraped off in order to ascertain the fineness of the metal. In most objects this scraping mark is rubbed down and cleaned off in the final polishing ; but in many cases, where the scraping would not show, it has been left just as the piece came from the assay office. In some instances an assay may have been made at a date much later than that of manufacture, as the owner may have been uncertain as to the quality of the silver. If a person wished to purchase or sell a piece of silver, and was in doubt as to its being of standard quality, he could ascertain this by taking it to the assay office and getting it assayed. There would be no necessity to have it stamped if it was only to be turned into money. In former times silversmiths were notorious for trying to get hall-marked silver which was below standard. Numerous cases of fines having been imposed on them for this offence occur in the records of the Dublin Gold- smiths’ Guild. In a very large number of cases pieces of Dublin plate of the eighteenth century bear no date letter. This the Provost accounts for by the post- assaying process. For example, a piece of plate made early in the eighteenth century, and not assayed and stamped until, say, 1760, could not have a date-letter struck on it, as the exact year in which it was made was not known; therefore no letter at all was struck, There are, however, examples among the College plate, of supposed post-assayed pieces, which haye the date-letter, including a soup tureen, with inscribed date 1722, and hall- marks, with date-letter for 1781, and three cups inscribed, respectively, 76 Proceedings of the Royal Irish Academy. 1699, 1745, and 1751, and bearing date-letters for 1814, 1769, and 1817 | respectively. I am of opinion that the absence of the date-letter is due simply to care- lessness. As long as the crowned harp, Hibernia, and maker’s mark were struck, the assay-master, or official who actually struck the marks, thought that the date-letter did not matter. During the seventeenth and eighteenth centuries each punch was struck separately, but in the nineteenth century what are called press-punches were introduced. The press-punch contained in one piece all the hall-marks to be struck. Instances of careless marking often occur on plate. For instance, on old Cork plate the proper mark was the ship between two castles, generally in three separate punches; but pieces are often found with the maker’s mark struck instead of the ship, to save the trouble of taking up the ship punch. I have come across hundreds of pieces of Dublin plate, and also many of English plate, of the eighteenth century, which have no date-letter, but the hall-marks are contemporary with the date of manufacture, as indicated by the form and decoration. In former times, in plate, one style of an object generally lasted for a certain period, until another fashion arose, though, of course, there was a little overlapping in the styles. Now, however, you can find in a silver- smith’s shop, and made at the present day, pieces of plate in all styles, from Roman times to A.D. 1917. In old silver the style and decoration of the object generally tell you approximately the period in which it was made; but the collector of the distant future will not be able to rely on this evidence. The Provost states that he cannot detect any variation in the form of the crowned harp punch at different periods. To the casual observer the small crowned harps, such as are found on Dublin silver, may seem all alike, but on examining them closely a great difference will be apparent. From having examined immense quantities of old plate I could generally tell the approximate date of a piece of Dublin silver from the crowned harp alone, without the aid of a date-letter. There were, naturally, several die-sinkers employed, who from time to time cut the punches, and each had his own way of cutting them. The number of strings to the harp and the dots on the crown vary considerably, to say nothing of the general outline of the whole stamp. There is another and very serious flaw in the Provost’s theory of post-assay, and that is, that dates of gift are inscribed on pieces of plate which were never made or thought of at the period denoted by the inscribed Wesrropp—Dates on Plate belonging to Trinity College. 77 date. For example, you may find a teapot or coffee-pot with inscribed date, say, of 1560, and hall-marks of, say, 1750, or a match-box inseribed, 1750, and hall-marked 1900, and say that these articles were made at the dates inscribed on them. The only drawback to this assumption is that these particular objects were unheard of at the earlier dates. There are almost equally ridiculous instances among the plate of Trinity College, Dublin, if we rely on inscribed dates. There is an item in the College accounts I would like to draw attention to, and that is the quantity of plate purchased by the College from private owners and silversmiths during the eighteenth century. From 1758 to 1781 there was over £1800 worth of plate bought, including £175 paid to Robert Hopkins, a Dublin silversmith, in 1758; £659 to Lord Mornington in 1759; £350 to Mrs. Stone in 1765; and £474 to James Warren, a Dublin silver- smith, in 1775. Taking seven shillings an ounce as a fair average of the price of silver in the eighteenth century, this would amount to roughly 5200 oz. Of this plate, that bought from private owners would possibly have had a family crest or coat-of-arms engraved on it, and that bought from a silversmith would probably have been unerested. Now almost all the pieces of plate belonging to Trinity College, Dublin, with the exception of some spoons and forks, and a set of dishes and plates totalling roughly about 2000 oz., have an inscription setting forth the donor’s name and date, together with his arms. There is no entry, so far as I can trace, in the College books of any sale of plate by the College during the eighteenth century. My theory, which is of course open to correction, is that this silver was obtained, and in many cases inscribed with donors’ names and dates, in place of pieces which had become broken and unsightly, or pieces for which money had been given, but not actually purchased. If this theory is not correct, then what has become of about 3000 oz. of plate bought by the College ? I am not including in this 3000 oz. uninscribed knives, forks, and spoons, as almost all of these, amounting to about fifty-five dozen, are of nineteenth- century make. I shall now mention a few of the more important pieces belonging to the College, and try to show why the inscribed dates are not to be relied on for dating the period of manufacture. . To begin with two-handled cups. There are three of these of exactly similar pattern, one inscribed with date of gift 1690, one 1699, and the third 1791 (this latter the Provost says should be 1701). Now these three cups are all struck with the punch of the same Dublin silversmith, Joseph Jackson, and bear the Dublin hall-marks for 78 Proceedings of the Royal Irish Academy. about 1780-90. (The College paid £84 16s. 84d. to a silversmith in 1781 for plate and engraving.) Although similar in some ways to cups of the eurlier periods, the work of these three cups does not appear to be of those periods, and there is no sign of the handles having been replaced, as the Provost suggests. The handles are typical of the period of the hall-marks. It seems curious that three cups, given at different periods and by different donors, should all bear the same maker’s mark and hall-marks of a much later period. I am inclined to think that these three cups were made about 1780 or 1790 to replace the pieces given at much earlier dates, and which had become battered. Another cup, inscribed with date of gift 1751, bears the Dublin hall-marks and date-letter for 1817, and maker’s mark of James Lebas. ‘The cup itself is in the style of about 1750-1760, a plain-belted cup, but the hall-marks clearly show that it must have been made in 1817. If the piece was made in 1751, how could the date-letter for 1817 be put on, as the Provost contends, or how could the King’s-head punch, which did not come on until 1807, be struck? This also appears to be a piece made at a later period in the style of the date of gift. The silversmith could easily have procured a cup of the period as a model. A cup, inseribed with date of gift, 1699, bears the Dublin hall-marks and date-letter for 1814, and maker’s mark of James Lebas. This cup is of a style in fashion neither in 1699 nor in 1814, but belongs to about the middle of the eighteenth century. Made by the same silversmith as the last, he probably copied a mid-eighteenth-century model in both cases. If the cup was made in 1699, how could it have the duty-mark not struck until 1730, the King’s head not struck until 1807, and the date-letter for 1814? In any case a cup of this form was unknown in 1699. As the date-letter punches and generally the other punches were defaced annually, there could be no possibility of obtaining an old punch to mark plate with. Turning to salvers, there are several anomalous examples. A salver with inseribed date of gift 1714, bears the Dublin hall-marks for about 1750-60, but no date-letter, and maker's mark of Robert Hopkins. Robert Hopkins, according to the College accounts, was paid £148 in 1758 for plate for the College. This style of salver was not in fashion in 1714, the usual salver of that period being a plain circular one on a central foot, while this one has a shaped border and rests on four scroll feet. Salvers of the early part of the eighteenth century are found on four feet which are usually perpendicular in form ; but the shape of the salver is not circular, but more or less rectangular, with rounded or set back corners. This salver could not have been made in 1714, and it is curious that the hall-marks (including Hibernia, which could Wrsrropp— Dates on Plate helonging to Trinity Colleye. 79 not have been put on in 1714) agree with the date at which Robert Ifopkins supplied plate to the College. A salver with inscription and date 1692 bears the Dublin hall-marks for about 1760. The salver is circular, with shell and gadroon border, and rests on four ball-and-claw feet, Such a pattern was never heard of in 1692, the shell and gadroon edge not appearing for about fifty years later, The,same remarks apply to a salver with inscribed date 1693, and bearing the Dublin hall-marks for about 1760. There ave six other circular salvers, each on three feet, with inseribed dates varying from 1693 to 1702, and bearing hall-marks of about 1760 and 1730. There are four oblong sauce-boats and covers, with gadroon edges, each on four scroll feet, inscribed with dates of gift (different donors) of 1699 and 1710, and all bear the same Dublin hall-marks of about 1770 and the maker's mark, probably of John Locker. This form of sauce-boat was unknown at the dates indicated by the inscriptions. Sauce-boats are found dating from about the second decade of the eighteenth century, the earliest ones being somewhat buat-shaped, with a lip at either end, and two handles, one on each side. Two saltcellars, inscribed with date of gift 1707, bear the Dublin hall- marks for about 1760 and maker’s mark of probably James Warren. These saltcellars are circular, each on three lion head feet—a fashion which did not come in until about 1720-30, The saltcellars of 1707 would have been the ovdinary trencher-salt pattern, which lasted down till about 1725 or even later. Pieces of plate such as wine-coolers, boxes, &c., having four claw or seroll feet, are found dating from the second half of the seventeenth century; and standing salts, tankards, &c., with three ball feet are found of the latter part of the sixteenth century; but the three feet with lion or human masks or shell besses, found on cream-ewers, sauce-boats, saltcellars, &c., were not introduced until between 1720 and 1730. It is a curious coincidence that a number of pieces of the plate belonging to Trinity College, Dublin, bear the maker’s mark IW, probably that of James Warren, a Dublin silversmith, working from about 1750 to 1790, and that in 1775 James Warren, silver- smith, was paid the sum of £474, presumably for plate supplied to the College. There are also belonging to the College forty-four pistol-handle knives and forty-eight pistol-handle forks, with inscribed dates varying from 1708 to 1/30, and all bearing the same Dublin hall-marks for about 1780-40. Tt seems curious that these pieces given at such different dates and by different donors should all bear identically the same hall-mark of a later period. R.I.A, PROC., VOL. XXXIV, SEOT. CG. [13 | SO Proceedings of the Royal Trish Academy. The same remark applies to eight baluster-stem candlesticks, given at different dates from 1694 to 1722, and by different donors, but which all bear the same Dublin hall-marks for about 1745. The College possesses several pieces of plate with inscribed dates, given late in the nineteenth century and early in the twentieth, but with hall- marks of some fifty or sixty years earlier, If hall-marks are not to be relied on, then why not say with analogous reasoning that these pieces were made at the date set forth in the inscription ? These examples are, I think, sufficient, although there are others among the College plate, to show that the inscribed dates cannot be relied on for proving the date of manufacture. As I have pointed out, many of the pieces could not have been made at the earlier lates, as the form of the particular object was unknown at the period. None of the pieces shows the slightest trace of any earlier hall- or makers’- marks, With all due deference to the Provost, I fail to see that his theory shows that the inseribed dates on the College plate, or on any other plate, prove the date of manufacture. AS) VL. ASSOCIATED FINDS OF IRISH NEOLITHIC CELTS. By E. C. R. ARMSTRONG, M.R.1.A., F.S.A. Read Novemner 30, 1917. Published Fesrvuary 21, 1918. No apology is needed for placing before the Academy the following short account of some finds of Neolithic celts. Finds of associated implements and weapons belonging to the Irish Bronze Age have often been published ; but those of the Neolithic Period have met with less consideration. It is, however, only by the publication and examination of associated finds that it will become possible to arrauge Neolithic antiquities in a progressive series. The Neolithic finds of other countries have been described from time to time.! An interesting recent contribution to the subject is that of Dr. Karl Schumacher,? who has published an illustrated account of the finds of associated implements of the Neolithic Period discovered in West Germany. The information about the finds included in the present paper has been obtained partly by searching the Museum Registers and Wilde’s Catalogue ; partly from information supplied by private collectors; and partly from accounts that have appeared in archaeological publications. Some remarks of a speculative character are added to the latter portion of the paper regarding the classification of the two principal forms of Neolithic polished axes, i.e. the type in which the outline is triangular and the butt pointed, and the type in which the outline assumes a more rect- angular shape, while the butt is rounded or roughly squared. Should such a classification be established by further researches, it will materially assist the arrangement of antiquities of the later portion of the Stone Age. The Collection of the Royal Irish Academy, preserved in the National Museum, Dublin, contains a large number of stone celts and chisels of various types. Unfortunately, in the early days of the collection, the records as to 1See Déchelette, Manuel d’archéologie, i, p. 516, note; Congrés International @ Anthropol. et Archéol. Préhistoriques, Gendve, 1912, ii, pp. 46, 47 ; Praehistorische Zeitschrift, ii, pp. 57-60; iii, p. 170; iv, pp. 281, 232; v, p. 262; and Mannus, v, 300. * Praehistorische Zeitschrift, vi, p. 29, sqq. R.I.A. PROC., VOL, XXXIV, SECT, 0. {14} 82 Proceedings of the Royal Irish Academy. the finding and associations of objects were frequently not preserved, so that these details are unknown about many of the specimens in the Museum. But so far as the records are available, they show that few celts have been found associated with burials:' the greater number have been accidentally discovered in agricultural operations, in draining or excavating the beds of rivers or the neighbourhood of lakes, or in cutting turf bogs for fuel. It is often difficult to determine whether associated finds of stone celts not obtained from interments, should be considered either as implement-- makers’ hoards, or as votive offerings, or as household goods hidden by their possessors in the ground when hurriedly compelled to leave their dwellings owing to some unfriendly raid or other cause ; in the latter case the recovery of the implements may have been prevented by the death of the owners or their flight to a remote locality. To ascertain the material of which a polished celt is composed is not easy. Geologists are shy about giving an opinion without examining a section of the implement under a microscope. Private collectors and curators of Museums sometimes have difficulty in getting their implements properly examined: thus statements as to the composition of celts are often inaccu- rate; and in many cases it has not been possible to give the composition of the celts mentioned in the course of the present paper. The finds are described in the alphabetical order of the counties in which they have occurred :— Co. Antrim.—Mr. W. J. Knowles, M.R.1.4., of Ballymena, Co. Antrim, whose collection of Irish Stone-Age antiquities is known to all students of the subject, possesses an interesting find of six polished stone celts. They were discovered together close to a sandstone grinding slab, four feet from the surface, in the brick-clay of Culbane. Mr. Knowles* has suggested that the person who ground the implements had intended to return the next day and grind others, but that in the meantime a flood had occurred in the Bann River, which either took a long time in subsiding, or else covered the axes and slab with a deposit of mud so that the place could not again be found. The objects comprised in this find have been lent to the Academy for exhibi- tion in the National Museum. The grinding slab has been illustrated by Mr. Knowles ;? it is made of sandstone, and measures 13 inches in length and 8 inches in breadth; it has been ground on both its flat faces. The butts 1 This also appears to be the case in Scotland; see Anderson, Scotland in Pagan Times (Bronze and Stone Ages), pp. 305, 306. 2 Proc. Royal Irish Academy, xxx, Sec. C, p. 219, 3 Jbid., Plate XTX, 126, Armsrronc—Associated Finds of Irish Neolithic Celts. 83 of all the celts are sharply pointed, except that of the smallest, which either has not been ground to so fine a point as the others or has been subsequently damaged. They do not vary much in size: the longest measures 52 inches in length, and the smallest 34% inches. Mr. Knowles also has in his collection two chipped celts which were found together at Rasharkin, Co. Antrim. One of these is made from flint. These two celts have also been lent to the Academy, and their owner has allowed them to be illustrated. (Fig. 1, 11 and 12.) Ti7 TA = Fie. 1. Associated finds of stone celts from various localities. (One-fourth.) In his paper on the Cushendall finds, Mr. Knowles' figured five chipped stone celts. Two are roughly blocked out; the others, which are finished, are of the same type: they were found together at Glenariff. He also has in his collection a large unground celt, fourteen inches in length, which was ' Journal Royal Anthropological Institute, xxxiii, p. 364, and Pl. XXXTIT,. (14%) 84 Proceedings of the Royal Irish Academy. found, with a slightly smaller specimen, “sticking with their edges in the ground,” in the townland of Knockans, near Cushendall.t To Mr. H. C. Lawlor, -R.1.4., I am indebted for an account of an inter- esting find of stone celts, discovered, about 1872, at Danesfort, Malone Road, Belfast, when making a small plantation. The find included nineteen stone axes, which are at present in Mr. Lawlor’s custody. ‘Three urns, described as of food-vessel type, were discovered, about fifty yards away from the celts, when making a drain. Figure 2, made from a photograph taken by Mr. W. A. Green, of Belfast, shows the celts reduced to about a ninth of their actual size. A twelve-inch scale is placed next to the largest celt; Stone celts found together at Danesfort, Malone Road, Belfast. From a photograph by Mr. W. A. Green. as can be seen from the illustration, they vary considerably in size and shape: some have pointed, others rounded, butts: one is chisel-shaped. Their exact lithological character has not been determined; but they are made from some hard rock, which is possibly the type of basalt found in parts of Co. Antrim. Journal Royal Society of A ntiquaries of Jreland, XXXVI, p- 386. Armsrronae—Associated Pinds of Irish Neolithic Celts. 85 A partly ground flint celt and a flint scraper were found, presumably together, at Dunboy, near Ballymoney, Co. Antrim. They were obtained by the Academy from the late Mr. S. F. Milligan, M.R.1.A. Co. Cavan.—A polished stone celt, with a square-shaped butt, two urns, only one of which has been preserved (a food-vessel of advanced form), a flint knife, a flint fabricator, and an object of bone, were discovered together in a cist in a tumulus at Killicarney, Co. Cavan. The objects were presented to the Academy in 1879 by the Earl of Enniskillen, through Mr. Loftus Tottenham, m.p. The urn, celt, and other objects have been figured and described by the late Mr. W. F. Wakeman.’ Co. Fermanagh.—A polished stone celt, with a rounded butt, made of amygdaloidal porphyrite, was found with burnt bones and charcoal in the south side of the carn, on 'l’opped Mountain, near Enniskillen, Co. Fermanagh.? Fra. 3. Stone celts found at Canrower, Oughterard, Co. Galway. (One-fourth.) Co. Galway.—Three stone celts were discovered under the root of a large deal tree, in the shallow bog of the townland of Canrower, near Oughterard, Co. Galway. They were found by a man named Naughton, (who), “Having dug round the root, he put his hand under it to raise it, and brought out these 2 Plunkett and Coffey, Proc. Royal Trish Academy, xx, pp. 656, 657. 3 Wilde, Catalogue of the Antiquities of Stone in the Musewm of the Royal Trish Academy, 1857, pp. 58, 59. 86 Proceedings of the Royal Irish Academy. greenish-grey felstone. They are highly polished. Their butt-ends appear to have been broken off; but the butt of one has been smoothed and polished, and a saltire has been cut upon it. These celts are interesting specimens: the cutting edges of two are slightly expanded, as if they were late in date, and had been reacted on by a metal form. It hardly appears likely that they were used to fell the tree, were broken in the attempt, and then thrown aside. Possibly they were deposited at the root of the tree as a votive offering. ‘lhe worship of trees was so widespread in prehistoric times' that such an offering need cause no surprise: while the axe was frequently regarded as a sacred symbol by peoples of antiquity; so that its votive use is easy to understand. Another celt, formerly in the collection of Dr. Petrie, has its butt-end, which seems to have been broken, polished and ornamented this specimen, and the celt above described, appear to be the only two in the Academy’s collection which show this curious feature. Two roughly made celts, now in the possession of Professor Lucas White » King, M.R.LA., are stated to have been found together at Menlough, Co. Galway. The butt of one is pointed; it has a flattened side. ‘The butt of the other is oval. Professor White King has allowed these celts to be illustrated. (Fig. 1, 9 and 10.) A small stone celt, 4 inches in length, and 143 inch in breadth, and portions of what were probably two others, are in the Academy’s collec- tion. All are stated to “have been found together, close to the castle of Aughnanure, near Oughterard, Co. Galway. The perfect celt has a butt-end closely approaching a point. It is not possible to determine exactly the shape of the broken specimens. ‘0, Kilkenny.—Two large polished stone celts, formerly in the Dawson collection, of similar shape, with roughly squared butts, are stated by Wilde? to have been found at Baysrath, Co. Kilkenny, The two celts are so alike in form as to lead to the supposition that they were fashioned by the same hand. There can be little doubt they were found in association. (Fig. 4.) Co. Limerick.—There are a number of stone celts in the Academy’s collection which are described as having been procured from Lough Gur, Co. Limerick. Lough Gur is a lake situated some three miles north of Bruff. Many antiquities of stone, bronze, and iron have been obtained either from the lake or near it, but as a rule particulars as to the exact spot where they were obtained are wanting. The lake appears to have been a place of ' See Frazer, The Golden Bough, 3rd edition, ii, Part 1, p. 7, sqq- 2 Op. cit., p. 51. Armsrrona— Associated Finds of Irish Neolithic Celts. 87 importance and sanctity in prehistoric times; objects being deposited in or near it probably as votive offerings to the water spirits." Two celts from Lough Gur are definitely stated to have been found together; they were purchased on 11 May, 1892, from J.J. Keane. One is a well-formed specimen with a narrow butt; the butt of the other is broken. (Fig. 1,5 and 6.) It is unfortunate that the exact spot where these two celts were found was not recorded: they are registered as having been found together in the lake. Five stone celts found at Lough Gur were purchased from W. Hinchy on 22 May, 1865. Two are flat and wedge-shaped, belonging to the type distinguished by Sir John Evans as having flat sides.* Another is a good Fic. 4. Stone celts found at Baysrath, Co. Kilkenny. (One-fourth.) specimen with a rounded butt; the butt of the fourth is pointed ; the fifth is a small, flat celt with a rounded butt. It is possible that these celts may have been in association when found; but the only information available is, as stated above, that they were purchased on the same date from the same person; they may have been discovered at different times. Three celts found at Lough Gur were purchased on 9 July, 1891, from the same person. One of these has a butt of intermediate type ; in another 1 For instances of similar customs see Déchelette, Manuel d’archéologie, ii, pp. 451- 453. 2 Stone Implements, 2nd edition, 1897, p. 98. 88 Proceedings of the Royal Irish Academy. the butt is unground; the third is irregular in shape, it is flat on one side and at the butt: that these celts were found in association is uncertain. Wilde' describes six stone celts as having been procured from Lough Gur, but he does not state whether they were found together. The remaining celts that came from this locality were procured for the collection at different times. Two stone celts were discovered, presumably together, when ploughing a field at Ballinacarriga, Co. Limerick; they were presented to the Academy by Colonel Gloster. Both are rectangular in shape ; the butt of one is square, that of the other is irregular. (Fig. 1, 7 and 8.) Co. Longford.—'T'wo stone celts, a flint fllake, and a tanged and barbed arrow-head, all purchased from Mr. Patrick Trapp, of Longford, are stated to have been found together in a field in the townland of Soran, parish of Clon- broney, Co, Longford, by Mr. John Sexton. The larger celt shows traces of use at the cutting edge; its butt appears to have been broken off. The cutting edge of the other celt also shows traces of use, its sides are ground, and it has the general appearance of being copied from a metal celt. The flake may have been used as a knife; both its edges appear to have been worked: the arrow-head is of the ordinary tanged and barbed type. (Fig. 1, 1, 2, 3, and 4.) Co. Meath—Two polished stone celts were found at Knockmooney, Faganstown, near Navan, Co. Meath, in 1843; they were presented to the Academy by Mr. J. R. Taaffe. One is oblong in shape, with a round butt; the other is rectangular ; its butt has been broken, Co. Monaghan.—Two polished green-stone celts were discovered with two vessels of baked clay on the property of Captain Stopford, at Lislea, near Clones, Co. Monaghan. The two celts, and one of the vessels, were given, in 1866, by Captain Stopford to the late Mr. Robert Day, of Cork. The vessel, a typically Neolithic urn with a round base, has unfortunately got separated from the celts ; it was purchased by the Academy after the sale of the Day collection in 1915. The two celts, as figured and described with the urn in 1904 by the late Mr. George Coffey, M.R.LA.,? were similar in appear- ance, and measured some 82 inches in length; their butts approach the pointed, more closely than the squared, type. It is desirable that their present habitat should be ascertained. ' Op. cit., p. 65. 2 Journal Royal Society of Antiquaries of Ireland, xxxiv, pp. 271-273. Armsrrong—Associaled Finds of Irish Neolithic Celts. 89 The Academy’s collection contains another Neolithic vessel ornamented with incised stripes, and having a rounded base; it was found in a subterranean cavern approached by a narrow passage, beside the moat of Dunagore, Co. Antrim. A stone celt and a number of flint arrow-heads are stated to have been found with the vessel, but they have not been preserved. A stone celt with a pointed butt, and a hammer stone, were found, presumably together, in the townland of Carn, Latnamara, Newbliss, Co. Monaghan: they were purchased by the Academy from Mr. Michael Croarkin. Co. Sligo.—The Academy’s collection contains a food-vessel of advanced type, catalogued by Wilde? as “found at Rathbarn, five miles west of Collooney, county of Sligo, in the sammit of an ancient rath, ‘in a square coffer of flag- stones, placed on edge, and contained burned bones and the small mica slate disc,’ which stands in front of it on the shelf.”” Col. W. G. Wood- Martin, M.R.1.A,, in a memoir on The Rude Stone Monuments of Ireland, published some thirty years later than Wilde’s Catalogue, figured this urn, the disc, and also a small celt of shale with a pointed butt, which he described as having been found with the urn and dise. Wilde! has catalogued this dise (W. 512) as “found near Rathbarn, Co. Sligo.” It seems doubtful if the celt was discovered in actual association with the urn and dise, for Wilde was usually careful to mention objects that had been found together; and the fact of his having done this in the case of the dise, but neglected the celt, leaves the association of the objects open to question: therefore it appears better not to use this find as evidence. Co. Westmeath.—Two stone celts were found together when ploughing a field at Clonrelick, Mount Temple, Co. Westmeath. They were purchased by the Academy in 1912. The butt of the larger celt has been broken off, and the end ground; that of the smaller has been slightly flattened. (Fig. 1, 13 and 14.) It is probable, from the name of the locality, Clonrelick, ze. the meadow of the cemetery, that these celts formed portion of the furniture of a grave, in which case the bones had possibly either perished or were not noticed by the person who found the celts. 1 Wilde, op. cit., p. 184. 2 Ibid., p. 189. 3 Journal Royal Society of Antiquaries of Ireland, xviii, pp. 271, 272. Abereromby (Bronze Age Pottery, i, pp. 126 and 143) follows Wood-Martin in describing the objects as found together. 4 Op. cit. p. 71. R.I.A. PROC., VOL, XXXIV, SECT. 0. [15] 90 Proceedings of the Royal Irish Academy. Two stone celts were found, in 1867, by Andrew Corrigan of Coolaleena in a bog on the property of Mr. Edward Stanley, in the townland of Muckanagh, Noughaval, Kilkenny West, Co. Westmeath. One of these is flat, with straight sides and a broad butt; the other is chisel-shaped. (Fig. 1, 15 and 16.) . Co. Wexford.—A find which, if genuine, is of interest as indicating the continued use of stone axes during the Bronze Age, is stated to have been discovered in 1892 by William Barrett, who found a stone celt associated with a bronze socketed celt at a depth of six feet in a bog at Ballyday, Co. Wexford. These objects were acquired for the Academy’s collection in 1914 from John Troy: the stone celt, which has a roughly squared butt, is made of slaty stone; it measures 35%; inches in length and 22 inches across above the cutting edge. Some mention must be made of the best-known sites where celts or implements were manufactured, and where consequently they have been found in more or less close association. The geology and archwology of Larne, in Co. Antrim, the most important of these, have been studied and described by Mr. R. Lloyd Praeger in conjunc- tion with the late Mr. George Coffey.'’ The typical whitish and much-rolled Larne flakes are too well known to require further description. ‘he so-called “Larne Celt” is of importance: a number of these implements are figured by Coffey and Praeger? Coffey, who placed the industry in the earlier stages of the Neolithic Period, considered the Larne celt to be a roughed-out stage in the manufacture of Kitchen-midden axes. But the extreme similarity between the Larne celt and the Campigny pick cannot be ignored; it is possible that the Larne celt is a finished implement, the whole culture being of the same period as that of Campigny. If this be so, the question of date presents a difficulty, for it is uncertain whether the culture of Campigny is to be regarded as a distinct chronological division of the Stoue Age or merely as a local manifestation. On this point it is desirable to bear in mind the cautious view taken about this industry by the late M. Déchelette.* Several sites are known in Co. Antrim where stone implements were made, the most important being near Cushendall, where Mr. W. J. Knowles, M.R.1.A., who discovered the site, obtained 1,812 unground celts, 273 broken specimens, and 240 ground and polished celts, making in all over ’ Coffey and Praeger, Proc. Royal Trish Academy, xxv, pp. 143-200, * [bid., p. 183. * Op. cit., i, pp. 326-829, Armstrong—A ssociated Finds of Irish Neolithic Celts. 91 2,000 axes from Glen Ballyemon and Tievebulliagh. In addition to these, others were obtained by collectors who acquired flakes and celts after the sites became known. Mr. Knowles, who presented a series of implements from this important find to the National Collection, has described the site, and illustrated the various types of implements he collected.’ The celts, &c., are made of black rock, probably an altered diorite, and are chipped, not polished. As well as the celts, picks, dises, chopper-like implements, rounded hammer-stones of black rock, and larger ones made of quartzite boulders, were discovered, together with thousands of flakes. Mr. Knowles considers the celts to have been roughed out on this site, and then carried away to other places to be ground and polished. The age of the implements is not clear; but Mr. Knowles? wrote: “The rude axes from Tievebulliagh and Ballyemon being found below the peat, and even mixed with the clay on which it rests, is, I think, satisfactory proof that they are of the earliest date and belong to a very early stage in the neolithic period.” Mr. R. A. Smith, r.s.4., who has devoted considerable study to Stone-Age antiquities, refers in his paper On the Date of Grime’s Graves and Cissbury Flint-mines, to the Cushendall implements, certain of which he compares to those from Cissbury, while in others he recognizes the hand-axe of the Drifc ; the side-seraper of le Moustier; the edge-trimming of Aurignac ; and the culture of Campigny. Mr. Smith considers that the culture represented by the remains at Grime’s Graves and Cissbury is contemporary with the Paleolithic Cave-period. If this be so, it would seem that the Cushendall implements are also to be referred to the earlier Stone Age, and thus do not come within the scope of the present discussion. A large number of stone celts have been found in the Shannon Fords. Wilde* wrote in 1857 that “The Academy is indebted to a Commission appointed for deepening and improving the navigation of the river Shannon, for the acquisition of more than one-half of the stone celts in the Collection.” The celts mentioned by Wilde were presented to the Academy on behalf of the Shannon Commissioners on 9 January, 1843, by Mr. Griffith, who stated that the celts were found at the fords of Keelogue and Meelick on the Shannon. These are the first points on the river passable except by boat ' Jounal of the Royal Anthropological Institwte, xxxiii, pp. 360-366; and Journal Royal Society of Antiquaries of Ireland, xxxvi, pp. 388-394. 2 Jowrnal Royal Anthropological Institute, xxxiii, p. 366, ° Archaeologia, lxiii, p. 141. + Op. cit., p. 48. ® Proc. Royal Irish Academy, ii, pp. 312-516. po" 92 Proceedings of the Royal Irish Academy. - above the falls at Killaloe, and are the main passes between the counties of Galway and King’s, and those of Clare and Tipperary. To improve the river for navigation it was necessary to deepen its bed at Keelogue ford by excavating to a depth of six feet below the bottom. A portion of the river was dammed off, one hundred feet in width and seven hundred feet in length. ‘The material excavated consisted of two feet of gravel, loose stones, and sand at the top; and of four feet of a mass composed of indurated clay and rolled limestone at the bottom. The loose material at the first two feet yielded a number of bronze swords, spears, &c., while the stone celts were found towards the lower part of the upper two feet. The greater number of the celts were stated to be made from a siliceous rock known as Lydian stone which occurs in thin beds stratified between the impure limestone called Calp, and is abundant in the neighbourhood of Keelogue and Banagher, but others were said to be made from a suberystalline and apparently igneous porphyritic rock, none of which occurs in the neighbourhood. Formerly it was thought that the discovery of such a large number of weapons at the river-passes indicated battles which had occurred between the men of Leinster and Connacht when disputing possession of the fords. But, as many of the celts are of the adze type, it is probable that they are to be regarded as implements rather than weapons: possibly they were used in making dug-out canoes. The few celts which are made of harder stone, and are more perfect in shape, may have been weapons. It is, however, to be remembered that only the shallow portions of the river were examined ; the contents of the deeper parts are unknown. More than three hundred celts of all types from the Shannon fords are in the Academy’s collection. Eighty-seven of these were not acquired by the Museum until 1911, when they were presented by Mr. T. P. S. Crosthwaite, who described them as found in the Shannon River, at Killaloe, when the bed of the river was being deepened some three or four feet by excavation and dredging. Most of them were found above the bridge; but some were obtained below it, while two were obtained under it, at almost the third’arch from the Co. Clare side. The shapes of the celts from the Shannon vary considerably ; but it is not possible to draw any conclusions from this, as it is uncertain whether they were deposited together. The finds of celts having been described, the theoretical portion of the paper may now be discussed; 7.¢., whether the celt with a triangular outline and pointed butt is an earlier form than that in which the shape is more rectangular and the butt either rounded or roughly squared. Armustrrona —Associated Finds of Irish Neolithic Celts. 93 The Neolithic culture of Scandinavia has been divided into four periods, the first three having a characteristic type of stone celt.1 In the earliest period the celts are made of flint, and are pointed-oval in section. Some are polished. In the second period the celts have squared sides and a thin butt; while in the third and fourth periods the sides of the celt are squared and the butt is broad. Some attempts have been made in other countries to determine the com- parative age of the different types; and describing, in 1912, five triangular stone celts, with pointed butts, found together at Bussleben, Thuringen, H. Motefindt* wrote that the question of the attribution of individual forms of stone celts to certain divisions of the Neolithic Period claims our interest in increasing measure. The solution of the matter seems to centre upon the position occupied by the triangular-shaped celt with a pointed butt. Schuchhardt* considers that in France the celt with a pointed butt was evolved from the form in use in the last stage of the Paleolithic Period, as shown by the chipped, unpolished implements of the culture of La Madeleine and that of Campigny. Schumacher‘ agrees with this view. He considers such celts to have originated in Western Europe, especially in France. Although Déchelette*’ distinguished three French types of celts, which he described as triangular, rectangular, and cylindrical, he did not place them in a progressive series, but appeared to consider that the varieties of form and shape were due either to the use for which the celts were designed, or to the natural form of the original stone from which the object was fashioned. It must be borne in mind, as Schumacher pointed out, that there is a connexion between the various shapes of the butts of stone celts and the manner in which they were hafted. The method of doing this varied in different countries ; in Switzerland, the basin of the Sadne, the basin of the Rhone, and the south of France, the most usual method was to fix the butt of the celt into either a kneed shaft, or a tenon of deer’s horn fixed into the wooden shaft; occasionally the head of the tenon was pierced and the wooden shaft passed through it. In afew cases at Robenhausen the stone celt was fixed directly into a club-shaped shaft. In England, north-west France, and ' British Museum Stone Age Guide, 1911, p. 99; also Déchelette, op. cit., i, p. 334; and Hoernes, Natw: und Urgeschichte des Menschen, ii, pp. 183-185. 2 Praehistorische Zeitschrift, iv, p. 231. 3 Sitzwngsberichte der Koniglich Preussischen Akademie der Wissenschaften, 1913 (2nd half year), p. 746. 4 Praehistorische Zeitschrift, vi, p. 43. 5 Op, cit., i, pp. 513, 515. 5 Op. cit., pp. 44-46. 94 Proceedings of the Royal Irish Academy. Belgium the celts, which are often long and pointed, were generally hafted directly into a wooden handle through an oval hole. That this was also the case in Ireland is indicated by the small stone celt with a pointed butt inserted directly into a wooden haft, which was found at Maguire’s Bridge, Co, Fermanagh, and is now preserved in the Academy’s collection in the National Museum, Dublin. A wooden handle for a celt, found at Coal bog, Boho, Co. Fermanagh, also in the same collection, has an oval hole in the head of the haft for the insertion of the celt. Sir John Evans' arranged polished celts in four classes, 7.e. :— (1) Those sharp or but slightly rounded at the sides, and presenting a puinted oval or vesica piscis in section. (2) Those with flat sides. (3) Those with an oval section. (4) Those presenting abnormal peculiarities. These divisions were adopted in order to describe celts more con- veniently ; it was not suggested that they denoted any difference in the date of the implements. Without going into a more minute classification, it is sufficient for the present to deal with two broad types of Irish Neolithic celts, 7.¢., those which are more or less triangular in outline and have a pointed butt; and those in which the outline assumes a rectangular form, and the butt is either considerably rounded or roughly squared. Something, no doubt, would depend upon the form from which the polished Neolithic celt was derived ; its origin has often been sought in the Danish kitchen-midden axe, and the late Mr. George Coffey wrote that “The kitchen-midden axe, or chisel, is typologically the beginning of the celt series.”* The kitchen-midden axe is a wedge-shaped implement, the cutting edge of which is formed by striking off a single large flake on each side. But Mr. R A. Smith has recently published a paper on Te Origin of the Neolithic Celt? in which he traces the evolution of one form of celt back through a large series to the “ point” of Le Moustier. The “point” of Le Moustier is roughly triangular in outline; so if Mr. Smith’s contention is correct, its shape may have exercised some influence on that of the Continental, and English, Neolithic celt with a pointed butt, which in turn may have influenced the Irish examples. ' Op. cit., p. 98. 2 Proc. Royal Irish Academy, xxv, Sec. C, p. 187. ® Archaeologia, \xvii, pp. 27-48. Armsrronc—Associated Finds of Irish Neolithic Celts. 93 The Academy’s collection contains about fourteen hundred polished stone celts and chisels, not including those found in the Shannon fords. Less than a quarter of these have pointed butts; they are not confined to any particular locality, so they cannot be considered as a local manifestation ; the remaining celts have rounded butts, or belong to intermediate types. A small number are flat in section and have their cutting edges expanded into a typical metal form; these are considered to have been made under the influence of metal celts, and to belong to the transitional period when metal implements were gradually replacing those of stone. Such celts can probably be considered on typological grounds as the latest form assumed by the stone celt. The evidence of the associated finds does not definitely settle the question as to the priority of the type with the pointed butt over that with the rounded butt. In some cases celts belonging to the two different types were found in association ; against this, however, must be placed the evidence of the larger number of finds which only included celts of similar type. The two celts found at Lislea have butts of the pointed type, and from their association with a Neolithic form of urn may be placed early in the series. On the other hand, the stone celt found with a food vessel of advanced form at Killicarney, Co. Cavan ;! that said to have been found with a bronze socketed celt at Ballyday, Co. Wexford, and the celt found with burnt bones in the carn on Topped Mountain, have rounded butts. In these cases the associations point to a late date for the celts, so that the evidence, such as it is, is on the whole in favour of what might have been expected on the analogy of other countries, 2.¢., that the celt with a rounded butt is a later form than that with the pointed butt. It might, therefore, be tentatively suggested that Irish celts belonging to the latter portion of the Neolithic Period could be arranged in three classes—those with a pointed butt and a triangular outline being regarded as the earliest ; those which are of a rectangular shape and have a much-rounded or square-shaped butt being looked upon as a later type; while celts which are flat in section, and have slichtly expanding cutting edges, would be con- sidered as having been made under the influence of metallic forms and be placed latest in the series. 1 But see ante, p. 89. If the celt with the pointed butt was really found in association with the urn at Rathbarn (Rathbarron), it would negative the above piece of evidence. R.A. PROC., VOL. XXXIV, SECT. C. [16] meee VII. THE DOMNACH AIRGID. By E. C. R, ARMSTRONG, F.S.A., M.R.1LA., AND THE Rev. PROFESSOR H. J. LAWLOR, D.D., Lirt.D. Puates III-V. Read Decemper 10,1917. Published Marcu 12, 1918. I. Descrirtion. By E. C. R. ArMsrrone. THE Domnach Airgid was obtained by the Academy in 1847; it had been previously described by Dr. G. Petrie, in the Academy’s Zvansactions, vol. xviii,! 1835-38 ; the Rey. Dr. J. H. Bernard? (now Archbishop of Dublin, and President of the Academy) also referred to the Domnach in his paper on the ms. found inside it; and a short account of the shrine extracted from Petrie’s paper was included in the Academy’s Celtic Christian Guide? Carleton’s Traits and Stories of the Irish Peasantry, ninth edition, 1869, vol. i, contains, pp. 402-5, two letters about the Domnach; the first of these, dated 15 August, 1832, was written by Dr. A. O’Beirne, who mentioned a drawing of the shrine apparently sent to Mr. Carleton, and added “In the chasing [sic] you will see they [the inscriptions] are referred to their supposed places.” The second letter, from Sir William Betham, is dated October, 1832; he wrote, “I have seen the Dona, which was exhibited at the last meeting of the Royal Irish Academy. It has been put together at a guess, but different from the drawing.” This drawing is probably the one referred to by Dr. O’Beirne. In the same letter Sir William Betham also wrote: “There are three inscriptions on the Dona: one on a scroll from the hand of the figure of the Baptist, of Ecck AcGNus Der. The two others are on plates of silver, but their exact position on the box is not marked in the drawing, but may be guessed by certain places which the plates exactly fit.” Another description of the Domnach might be considered unnecessary ; but Petrie’s paper, upon which all subsequent accounts have been based, was mainly directed to proving that the reliquary was the identical one given by ' Pp. 14, seq. ? Transactions Royal Irish Academy, xxx, p. 303, seq. 5 Pp, 45, 46, 94, 95, Armsrrona ann LAwLor—TVhe Domnach Airgid. 97 St. Patrick to St. Mac Cairthinn: it was published nearly eighty years ago ; and in view of the fresh evidence discovered by Dr. Lawlor, it has been considered desirable to re-describe and re-examine the shrine. Petrie! stated that recently someone had been employed to repair the case, and that certain of the ornamental plates had been replaced in an order different from their original one. The illustrations which accompanied his paper included views of the front, top, sides, and back, of the shrine; they were made from drawings by Mr. George Du Noyer, and show the mediaeval plates on the front and base differently arranged from their present position, Petrie* stated that the figures were ‘“ restored to their proper places in the accompanying plates, on the authority of Sir W. Betham’s drawings.” From the letters printed by Carleton it would seem that these included the drawing referred to by Dr. O’Beirne; whether they represented the Dommnach as it originally was, or are merely based upon the artist’s suppositions, is impossible to determine. The panel drawn by Du Noyer as occupying the first quarter of the front of the shrine is now in the fourth; his second panel is in its present position ; his third is now in the first quarter; while his fourth panel is the third of the present reconstruction. The small reversed inscription on the top edge of his fourth panel is omitted, while the large one on the upper rim is shown detached from the shrine. The drawing of the top of the shrine shows a crystal in the sinister setting, which is now empty: the sides are ornamented with the plates at present fastened to the base, St. Katherine being on the sinister, and the two others on the dexter side; the rim is also differently arranged. The drawing of the back omits the ornamentation on the upper limb of the cross and the inscription on the cross-piece. A small sketch of the Domnach, stated to have been drawn by Petrie, is appended to Carleton’s notes about the shrine; in this, the panel with St. Michael and the dragon is placed in the fourth quarter, and that with the bishop handing the case to an ecclesiastic in the third, In the short account of the shrine, printed in the Academy’s Celtie Christian Guide, the front and top were illustrated from photographs : it is proposed to include these two views in the present paper, and to illustrate the other parts of the case from photographs taken by Mr. A. Me Googan, and made use of by the permission of the Acting-Director of the National Museum, Dr. R. F. Scharff, M.R.I.A. To avoid confusion, the heraldic terms ‘‘ dexter” and “sinister” are used throughout the following description :—the dexter is the right side of the 1 Op. cit., p. 16. ° Op. cit., p. 16. [16*] 98 Proceedings of the Royal Irish Academy. shrine, which will be to the observer’s left; while to his right will be the sinister or left side. The shrine is badly preserved; this is easy to understand if Dr. O’Beirne’s' account of its adventures is true. Unfortunately, the repairs and restorations of more recent years have been roughly carried out; lumps of solder have been left exposed in prominent places, while the riveting on of the outer plates has been crudely executed. The Domnach Airgid is a box of yew wood, which was covered in the first instance with bronze panels coated over witha white metal. Petrie? stated that these panels were plated with silver; but I was doubtful if any process of silver-plating was in use in Ireland at such an early date, and so asked Dr. A. G. G. Leonard, of the Royal College of Science for Ireland, to determine the metal. This he did by means of a spectroscopic analysis. The first examination showed some faint lines of silver, which Dr. Leonard thought were due to impurities in the bronze; he therefore made a second examina- tion of the deposit, and has reported as follows : — “The result of the spectroscopic examination of the Domnach Airgid shows the deposit to be composed mainly of tin. The silver lines were not appreciably strengthened when the deposit was sparked on to the lines of the bronze, whereas the tin and some copper lines were much intensified.” It appears, therefore, that the bronze plates were coated with tin,’ not plated with silver. From the colour of this tin coating the shrine received its name of Domnach Airgid, iz, the silver Domnach. In the fourteenth century a further covering of silver-gilt plates, ornamented with figures of saints in relief, &c., was added, a rim of ornamental panels, one having an inscription, being placed round the case to cover the edges; it will be shown that there is reason to believe that the present back, and the embossed plate on the top of the shrine, were still later additions. The Domnach is now made up of the following parts :-— (1) The front, ornamented with the fourteenth-century crucifix and panels, which are directly attached to the inner framework, all traces of any earlier metal plate having apparently disappeared. Including the rim, it measures 230 mm. by 167 mm. ' Carleton, op. cit., p. 402. ? Op. cit., p. 15. * The practice of coating objects with white metal goes back to early times. Several Irish brooches, probably of seventh century date, in the Academy's collection, are so coated on their backs. The Lough Erne shrine, which has been assigned to the ninth century, is made of a yew-wood box covered with bronze panels which are ‘‘ apparently tinned.” Several objects of Scandinavian origin found in Ireland are coated in the same way. See Coffey, R. I. A. Celtic Christian Guide, 1910, pp. 23, 24, and 42-44. ARMSTRONG AND LawLor—The Domnach Airgid. 99 (2) The top, now covered by an embossed silver plate, it measures, including the rim, 226 mm. by 101 mm, (3) The sides. These are bronze panels coated with tin, and form part of the earliest metal covering of the shrine. Including the rim, they measure respectively 173 mm. by 92 mm.; and 175 mm. by 96 mm. (4) The base, now covered by three silver-gilt fourteenth-century panels, which are apparently riveted directly to the inner framework. Including the rim, it measures 228 mm. by 99 mm. (5) The hack: a plate of bronze or latten nailed to the inner wooden case; attached to it by rivets is an ornamented and inscribed cross of copper, gilt. Including the rim, the back measures 222 mm. by 167 mm. (6) The rim: portions of this have disappeared; the vacant spaces have been filled with pieces of brass. (7) An inner box of yew wood, which is described post, p. 105. The front of the case (Plate IIL) is thus ornamented: the centre is treated as a cross dividing the remaining space into four panels, each contained in a frame of niello work; the dexter lower panel has on its upper side an inscription, reversed, in raised Lombardic characters, which reads :— JOHANEAS : O BARRDAN: FABRIGAVIT Each panel measures, including the frame, roughly 89 mm.by 70mm, A figure of our Lord, of silvez, gilt, is attached to the cross; over His head is a bird, presumably a dove, ina setting of blue champlevé enamel ; above this is a small square silver setting, composed of a beaded base supporting a band of dots arranged in groups of five, with fleurs-de-lis claws, containing a crystal. Petrie! wrote that this setting probably contained “a supposed piece of the ? true cross.” Examination has shown this conjecture to have been correct ; see post, p. 104. Adjoining this setting, in a rectangular frame of blue enamel, is a heater-shaped shield bearing emblems of the Passion, the crown of thorns, three nails, and two hands holding scourges; remains of red champlevé enamel can be seen on the shield. The first of the silver-gilt panels is divided into two niches ; the dexter contains a figure of St. Michael and the dragon; the saint, who carries a shield emblazoned with a cross, stands upon the reptile and thrusts a cross-headed spear into its mouth; the sinister niche contains an effigy of the Virgin, crowned and enthroned, giving her breast to the child; she holds some small object in her left hand. The second panel is divided by niches into three compartments. In the first is a figure, wearing a round hat, dressed in a long robe; he holds in each hand a long-handled rod with a foliated top. ' Gp. cit., p. 1b. 100 Proceedings of the Royal Irish Academy. Sir William Betham in his letter, previously referred to, described this figure as a “ bishop pierced with two arrows.” Petrie* names it St. James. If so, possibly he is holding fuller’s teazles ; in the second compartment is an effigy of St. Peter with the two keys in his right hand, and in his left a book; the third niche contains a figure of St. Paul, holding in his right hand a sword, and a book in his left. The third panel has two compartments; the dexter contains, under a double arch, an effigy of a seated bishop, wearing a mitre, amice, cope, and albe, handing a box to an ecclesiastic. Petrie? suggested that this scene represented St. Patrick handing the Domnach Airgid to St. Mac Cairthinn. A female figure, wearing a hood and flowing robe, stands in the secondniche. The fourth panel is divided into three niches; the first contains an effigy, perhaps of St. Columba, dressed in a long robe, with his right hand on his breast, and clasping a book in his left; the figure of an abbess, holding in her right hand a floral-headed crozier and in her left a book, occupies the second niche; she is probably meant for St. Brigid ; in the third niche is depicted an archbishop, possibly St. Patrick, wearing a mitre, amice, apparelled albe, dalmatic, chasuble, and pall; his right hand is raised in benediction ; in his left is an archbishop’s cross. The top (Plate IIL) is covered by a silver plate, originally gilt, now broken into three parts. Its removal has shown that beneath it is a bronze panel, coated with tin, and decorated with interlaced knot-work of a character similar to that on the sides, but not bordered with fret-patterns ; the interlaced knots at the ends are more elaborately decorated than those in the centre (Plate V). The upper plate is ornamented with three round bosses. Each rises from a beaded rim, and measures some 65 mm. in diameter. The decoration of the dexter and sinister was similar. It con- sisted of an oblong crystal, contained in a fleurs-de-lis claw-setting, which formed the centre of a star with six rays, originally covered with blue champlevé enamel. Between the rays are alternate griffins and lions, looking backwards. The crystal on the sinister side is missing. The central boss contains a large oblong crystal in an elaborate setting of a beaded rim, from which rises a band, ornamented with dots grouped in fives, surmounted with fleurs-de-lis. The body of the boss is ornamented with a ribbon of beaded and linear work, arranged in a chevron pattern. ‘The upper of the triangular spaces made by the ribbon are filled with wire-work, each of the lower with two birds, pecking at a cross-headed staff. Between the central and outer bosses are four galloping horsemen. ‘They wear helmet-shaped hats, with wide brims, have large ruffles round their necks, and are clothed in a long * Carleton, op. cit., p. 404. 2 Op. cit., p. 15. 3 Thid., p. 18. Armsrron@ AND LawLtor—The Domnach Airgid. 101 pleated garment, belted at the waist. According to Irish custom they do not use stirrups. At two of the corners of the plate are small circular settings of beaded work, with fleurs-de-lis claws: the dexter is now empty ; the sinister contains a pearl: possibly these settings were originally attached to the corners of the rim. Petrie! stated that the crystals set into the bosses covered relics. Examination has failed to disclose these. The sides (Plate IV) are bronze panels, coated with tin, attached to the wooden box. They are engraved with interlaced work, bordered with a fret- pattern at the side edges. The base (Plate V) has, according to Petrie,’ lost its original ornaments, their place having been supplied by the “recent repairer” with panels taken from the sides: now it is covered by three silver-gilt panels, each roughly square, measuring about 70 mm. by 70 mm., ornamented with effigies in relief, contained in a beaded and linear framework. The dexter panel contains an effigy of St. John the Baptist, who wears a camel’s-hair robe, and holds a black enamelled disc, on which is the Agnus Dei, in his left hand; in his right is a scroll bearing an inscription in Lombardic characters aaade: ACNUS DAI. To the saint’s left, engraved on the field of the panel, is Salome, with the Baptist’s head on a charger. An effigy of St. Katherine crowned occupies the centre panel: she holds a wheel in her right hand, and in her left a book.. On her right is engraved a priest adoring her; and on her left is a server, swinging censers. In the sinister panel is a male effigy, enthroned, wearing a cope, fastened by a jewel, and an albe; his right hand is raised in benediction; he holds a small cross in his left: engraved at each side of him is a figure, swinging a censer. The back of the shrine (Plate LV) is a bronze or latten plate, attached by rivets to the wooden box, engraved with a plain ornament of straight lines round the edge, and with quarter-circles at the corners, and also round the centre of the copper-gilt cross which is riveted to the plate: the upright shaft of the cross is ornamented with a floral decoration; on its cross-piece is engraved an inscription in black letter. Petrie? wrote that he was unable wholly to decipher this inscription on account of ifs injured state, but that it ended with the word cloacbavr, the name of the see to which the reliquary had originally belonged.t After examining the inscription carefully I saw ' Op. cit., p. 16. 2 Ibid. 3 Op. cit., p. 17. ‘In the extract about the shrine printed in the Academy's Celtic Christian Guide, pp. 94, 95, Petrie’s reading of this inscription is repeated. In the preface to that work it is stated that Mr. R. I. Best, of the School of Irish Learning, had re-examined the inscriptions on the yarious shrines. Mr. Best is not, however, responsible for the repetition of this error: he only examined the Irish inscriptions on the different shrines. / 102 Proceedings of the Royal Irish Academy. that what Petrie had read as a ¢ was part of the symbol ihc which is placed on a small compartment of the inscription; on my pointing this out to Dr. Lawlor, he at once concurred. Professor R. A. 8. Macalister subse- quently examined the lettering with me, and, after working at it for a short time, we were able to decipher the names of the Magi, or Three Kings, the inscription reading :— jappar ib C melchicar ba ibe! ldarar The traditional names of the Magi, commonly known as the Three Kings of Cologne, are frequently found on mediaeval ornaments. As a magical formula they were supposed to be of especial efficacy against the falling sickness; they were also used as a charm against fever." , The rim (Plates III, [V, V) was composed of silver-gilt plates, apparently bent at right angles, so as to cover the edges of the case. The plates were joined lengthways by hinges, the longest centre-piece measuring 141 mm. in length. The corners were covered by three plates placed together at right-angles, their juncture being marked by a triangular-shaped ornament, decorated with a trefoil. ‘T’o the angles of this ornament were attached either three human, or three zoomorphic, heads, and to these were fastened small circular settings, with beaded bases, and fleurs-de-lis claws, now empty, with the exception of one on the dexter side, which con- tains the remains of a pearl. Much of the early rim has perished, and its place has been supplied with pieces of brass. The upper rim on the front (Plate III) is reversed. Its centre-piece contains, in niello work, an inscription, in Lombardic characters, which reads :— JOWS : O KARBRI : COMORBANUS : S: TIGNACII : PMISIT From this inseription the date of certain parts of the shrine can be ascer- tained, for the death of John Ua Cairbri, successor of St. Tigernach in Cluain-Eois, is recorded in the Annals of Ulster; at the year 1353. There can be no doubt that this is the ecclesiastic who is commemorated on the Domnach Airgid. ‘Yo the dexter side of the inscription is hinged a small silver panel, in a beaded frame, containing a hare at speed. A zoomorphic head divides this from a similar panel, containing a running hound, with a jewelled eye. On the sinister a panel of the same type, but containing a wyvern, with a floriated tail and jewelled eye, is hinged to the inscribed panel: it is divided by a human head from a similar panel, forming part of 1 King, Archaeological Journal, xxiv, pp. 233, 234 ; also see Archacologia, xxx, p. 400 ; and Dalton, British Museum Catalogue of Mediaeval Finger Rings, p. 140. 2 ii, p. 497, and note. » Armsrrone and Lawtor—TZhe Domnach Airgid. 103 the rim at the side. The long panel at the top of the shrine contains two birds, with long floriated tails, placed on each side of a small, square setting, with a beadwork base and fleurs-de-lis claws, which contains a crystal: hinged to this, on the dexter, is a panel containing a hound, a zoomorphic head divides this from another, containing a hare, which is joined to a small panel containing a six-leaved flower. On the sinister side a panel containing a hare joins the long one, and a human head separates it from a hound and a small flower panel. On the base of the front of the shrine the long panel contains two birds, placed at either side of the Saviour’s feet, similar to those in the long panel at the top of the shrine: hinged to this, on the dexter, are two wyvern panels, divided at the corner by a human head; and, on the sinister, are two panels, one with a hare, and the other with a hound, separated at the corner by a zoomorphic head. The under side of the long panel, which is on the base of the case (Plate V), is plain: at each side of it are hinged hare, hound, and floral panels, parted on the dexter by a human, and on the sinister by a zoomorphic head. The rim on the upper dexter side of the case (Plate IV) consists of two panels of interlaced work, divided by a zoomorphic head, and one flower panel. At the base of this side are two wyvern panels, which a human head separates, and a flower panel. The rim on the upper end of the sinister side (Plate IV) consists of two wyvern panels, parted by a human head, and a flower panel; and on the lower, of two interlaced panels, separated by a zoomorphic head; and a flower panel. At the back of the shrine (Plate IV) the corners are composed of bronze or latten panels, placed together like those on the front; these panels are plain, with slightly ornamented borders, their junction is marked by what seem to have been zoomorphic heads; but these are now so much worn that it is impossible to be certain on the point. ‘The remaining undescribed portions of the rim are made of brass, and appear to be modern additions. The interior of the shrine and the contents of the settings may now be described. Permission having been given by the Council of the Academy, with the assistance of Dr. Lawlor, I opened the shrine on Wednesday, 7 November, 1917. Mr. A. C. Forbes, r.H.A.s., Forestry Inspector to the Department of Agriculture, kindly attended for the purpose of determining the species of wood of which the inside box was made; Mr.J.J. Buckley, M.r.1.4., Keeper of the Art and Industrial Collections in the National Museum, was also present. I first raised the claws and removed the crystal from the small square setting on the front of the case which Petrie’ stated “probably contains Op. cit., p. 15. R.I.A. PROC., VOL. XXXIV, SECT. C. [17] 104 Proceedings of the Royal Irish Academy. : a supposed piece of the true cross’’; its under surface was fiat ; it rested upon another crystal with four facets and a flattened point, fastened down by some kind of plaster, in which were two small fragments of wood, one measuring about five millimetres in length, the other being about the size of a pin’s head, and some dust. The smaller fragment crumbled into dust. Dr. G. H. Pethybridge, w.R.1LA., Professor H. H. Dixon, sc.p., F.R.S., Professor H. J. Seymour, B.sc., and Mr. A. C. Forbes, F.H.A.s., kindly examined the larger piece under a microscope; its small size rendered any determination as to its character doubtful, so Dr. Pethybridge undertook a further more searching examination, and has kindly sent me the following note (with which Professor Dixon concurred), giving the result :— “The piece of wood, said to be a portion of the true Cross, is of light colour, about five millimetres long, three-quarters of a millimetre wide, and one- fifth of a millimetre thick. Itis most improbable that this piece of wood can have been a portion of a piece of timber as ordinarily understood. On the other hand, microscopic examination of it both in cross-section and longi- tudinally goes to show almost with certainty that it must have been derived from a young twig or branch of a tree or shrub. Portion of the medullary sheath of the twig is present with primary wood (protoxylem), and with some half-dozen rows or so of elements of secondary wood. The secondary wood present appears to consist of tracheids only, having simple pits; but it is quite possible that had the specimen been larger, portions of the secondary wood, including vessels, might have been present. In one part of the fragment a structure resembling a small bordered pit was seen. From the small amount of tissue necessarily present in such a mere fragment of wood the identification of the tree or shrub from which it was taken would be a matter of great difficulty, and one involving probably an inordinate amount of time.” It is probable that the fragment represents the piece of the true Cross which is mentioned in the Life of St. Mac Cairthinn' as one of the relics contained in the shrine. It is recorded in the Annals of Tigernach* (Rawl. B. 488) at the year A.D. 1123, that a piece of Christ’s Cross was brought to Ireland, and that some of it was given to King Toirdelbach Hua Conchobair, who enshrined it at Roscommon: so a relic of the true Cross was brought to Ireland in the early part of the twelfth century. The small crystal in the square setting on the rim was next removed; beneath it was some dust. 1 See the passage quoted by Dr. Lawlor in part ii of this paper, post, p. 110, ? Revue Celtique, xvili, p. 43, seq. Armstrona anp Lawtor—Zhe Domnach Airgid. 105 The added silver plate on the top was then lifted off; below it was one of the original metal plates of the case, ornamented with interlaced work. Beneath the central boss of the added plate was a mixture of dust and small fragments of wocd; there was nothing beneath the crystal. No remains were found beneath the dexter and sinister bosses, and there was nothing under the crystal remaining on the dexter side. Mr. J. N. Halbert, mn.1.a., Assistant in the Natural History Section of the National Museum, Dublin, kindly examined the material found beneath the various settings, and has reported as follows :— “Most of the material handed to me for examination was obtained from under the central boss on the top of the shrine; a much smaller quantity was found under the square setting on the rim, and in the setting which contained the part of the true cross. A minute examination of this material shows that it is mainly composed of dust grains, numerous wool-fibres, fine bristles, small pieces of a soft white wood, fragments of insects and spiders, the whole mass being held loosely together by the fine strands of spider webs. A few small fragments superficially resembling human hair are present, but these proved to be nothing more than dyed sheep’s wool similar to that used in homespun cloths. At one time there may have been a layer of cloth! between the boss and the top of the shrine. If not too tightly fitted, this layer would supply a suitable habitat for the caterpillars of small house moths, and indeed the cocoons of such moths, as well as the remains of the insects, are present in the debris. The material found under the other ‘settings’ resembles that of the central boss.” The removal of the front disclosed the interior of the case: it was empty (Plate V). The wooden box is hand-cut; it has every indication of age. Its dimensions are:—interior length, 193 mm.; breadth, 140 mm. ; depth, 74 mm. ; the average thickness of the wood is about 11mm. A piece measuring about 119mm. by 25mm. has been broken away from the upper portion of the base; it has been partially repaired by a thin slip of wood. The outside measurements of the box are 218 mm. in length and 164mm. in breadth. he lid, which is in a bad state of repair, and has been covered at the back by a brass plate measuring 192 mm. by 143 mm., measures 218 mm. in length and 143 mm. in breadth. Mr. A. C. Forbes, F.H.A.S., and Dr. G. H. Pethybridge, M.R.1.4., kindly examined small samples from the front and one side of the box, also a speci- men of the thin piece let into the base, in order to determine the species of 1 See on this point Appendix I of Dr. Lawlor’s memoir on the Cathach (Proc. Royal Trish Academy, xxxiii, sec. C, p. 394), where instances are iven of pieces of cloth being inserted as relics in Irish shrines. [17*] 106 Proceedings of the Royal Irish Academy. wood. They inform me it is yew; the small piece of wood, inset apparently in modern times, to repair the box, is oak. The box (Plate V) has the appearance of having been made with a sliding lid; the two sides show a rebate; one end is slightly lower than the other and levelled to the rebate; but though the present breadth of the lid would allow it to rest on the rebated sides, it is the full length of the box, and if slid in and placed against the higher end it would leave a portion protruding from the case. The date, or rather dates, of the various parts of the shrine can now be considered. ‘The three remaining plates of the first metal covering available for examination, z.c., the lower plate on the top and the two sides, are un- inscribed, so that their dating must rest upon artistic grounds; they are decorated with interlaced work: the sides have a border of fret patterns. These interlaced patterns closely resemble those of the Book of Durrow. The evidence collected by Dr. Lawlor,! and published in his Memoir on the Cathach, points to the seventh century as the probable date for the Book of Durrow. Professor W. M. Lindsay,’ in his palaeographical notes appended to Dr. Lawlor’s Memoir, says that the Ms. cannot be much older than 700 «.D., but seems inclined to put it little later. Professor R. A. 8. Macalister’ has, hae ever, sought to prove that the middle of the ninth century is a more probable period for the writing of the Ms. in question. Now the interlaced style was not in general use in Ireland before the seventh century,‘ and as the ornament of the panels of the Domnach Airgid shows no trace of La Téne survivals, it cannot, I think, be earlier than the seventh century, while it is probably as late as the eighth. It is impossible to date with certainty the wooden box to which the metal sides were attached; if its lid was originally a sliding one, the protruding end would have hindered the fastening on of one of the metal ends, but the metal plates presumably always had a framework to which they were fastened; so it is probable that the wooden box is at least as old as they are. The inscriptions on the front of the shrine show that it was redecorated not later than the middle of the fourteenth century, and to this period may be assigned the silver panels with figures of saints in relief and the earlier portions of the rim. The silver plate now covering the top is more difficult to date. Petrie® considered it to be contemporary with the added panels on the front and * Proc. Royal Irish Academy, xxxiii, sec. C, pp. 317-322. ? Appendix ii of the same paper, p. 406; see also the Rey. S. F. H. Robinson, Celtic IMuminative Art, 1908, p. xx. > Essays and Studies presented to William Ridgeway, 1913, p. 301. * See Coffey, Royal Irish Academy's Celtic Christian Guide, 1910, p. 8. 5 Op. cit., p. 16. Armsrrone AnD Lawtor—TVhe Domnach Airgid. 107 base, and described one of the four horsemen represented upon it as exhibiting “with minute accuracy the costume of the nobility in Ireland during the 14th century.” But the most noticeable feature about the horsemen’s costume is the ruffles round their necks, and on this account I am not inclined to believe the panel to be older than the sixteenth century —an opinion, I may add, that is strengthened by the inferior craftmanship of the plate, which, though based upon, lacks the strength and vigour of, early mediaeval work. The Rev. J. E. McKenna, M.R1.A.,! has suggested that the horsemen were _ placed on the Domnach as a compliment to the Maguires, as a section of that family had adopted at an early period a mounted knight, sword in hand, for their “insignia or arms.’? This seems a reasonable way of accounting for the presence of such warlike figures on a sacred object: if correct, it would strengthen my view as to the late date of the top, for heraldic devices do not appear to have been adopted by the Lish chiefs until late Tudor times.* Petrie stated that the inscription on the cross which is attached to the back of the shrine was later in date than those on the front: this is evident. On the front Lombardic characters are used, while the inscription on the back is in black letter. ‘The examination of a number of ecclesiastical seals * has shown that good Lombardic characters were used from the early part of the thirteenth to the middle of the fourteenth century, when they were replaced by black letter, which, at first bold, afterwards fine and close, remained in use until about 1500 a.p. The metal back of the shrine may, therefore, be as late as the fifteenth century ; it is certainly not older than the latter portion of the fourteenth. Thus the Domnach Airgid appears to have been first constructed as a simple wooden box of uncertain age, having as early as the seventh or eighth century a metal casing of bronze plates, coated with tin, engraved with interlaced patterns. Before 1353 the crucifix, silver-gilt panels, and rim were added. The fifteenth century probably saw the addition of the back and cross inscribed with the names of the Three Kings; while in the sixteenth century the embossed plate was attached to the top. The pieces of brass-work to be seen on the rim and on the cross on the front of the shrine were possibly added by the person Petrie* alludes to as the “recent repairer.” 1 Ulster Journal of Archaeology, vii, 1901, p. 121. * Such a device doubtless originated from a common type of equestrian seal bearing the effigy of the chief. Sir Bryan Maguire, created a peer in 1627, bore arms of Vert, a man in armour on horseback with a sword in his right hand. See Burke, Dormant and Extinct Peerage, 1883, p. 349. 3 Armstrong, Jowrnal Royal Society of Antiquaries of Ireland, xliii, p. 66, seq. 4See Sir W. St. John Hope, Proc. Society of Antiquaries of London, 2nd S., xi, p> 305. 5 Op, cit., p. 16. 108 Proceedings of the Royal Irish Academy. Il. History. By H. J. Lawtor. A tradition was current in the upper part of the county Fermanagh a century ago that the shrine then known among the peasantry of that district as “the Domnach” was presented by the Pope to one Donagh O’Hanlon, who had made a pilgrimage to Rome about 600 years earlier. This story of course cannot be true; and it is vain to speculate as to the element of fact which may lie behind it. Its value is not enhanced by the assertion of the person who told it to Dr. O’Beirne of Portora in 1832, that it rested on the authority of Sir James Ware.’ It is nevertheless not without interest, as we shall see. But Dr. O’Beirne’s informant went on to relate the wanderings of the Domnach in the seventeenth century, with no appeal to authority: and here, I conceive, we may accept his statements as at least not far from the truth. Dr. O’Beirne, it is true, describes them as “vague detail”; but I am rather inclined to note the absence of vagueness and the minuteness of detail as a guarantee of the general accuracy of the tradition, This is what he said: O'Hanlon deposited the shrine in the monastery of Aghalurcher ; but when this monastery was destroyed in Cromwell’s time, it was hidden at Lough Eye, between Tempo and Lisbellaw, and after the restoration of peace it was placed in a neighbouring chapel. In 1689 it was again concealed at Largy, “an old castle at Sir H. Brooke's deer-park,” from which it was dug up after the Boyne by a priest named Anthony Maguire. All the places mentioned are in Maguire’s country in the county of Fermanagh, and not far apart. After Anthony Maguire's death it was carried by his niece to Florence Court; but the Maguires insisted that it should be brought back, and this was done. Here the story ends, and we hear no more of the shrine till the beginning of the nineteenth century, when it was in the possession of an old woman at Brookeborough, close to Largy, and in the Parish of Aghavea, which adjoins Aghalurcher. She was a Maguire and lived in Maguire’s country; and she declared that the shrine, which was supposed to contain some of the Blessed Virgin's hair, had belonged to “the lord of Enniskillen” —the chieftain of the Maguires who was executed for complicity in the rebellion of 1641.4 From ‘I do not know any passage in Ware’s published writings which refers to the Domnach. 2The Rey. J. E. McKenna believes that there was no religious establishment in Aghalurcher, other than a parish church, in later centuries. But in Fermanagh traditions old parochial churches are often called monasteries. The church of Aghalurcher was in ruins in 1622. The introduction of Cromwell at this point of the story is therefore unhistorical. , 3 A very small townland, named Largy, adjoins Deerpark in the parish of Aghavea. ! J. Groves in W. Shaw Mason’s Parochial Survey, 1819, vol. iii, p. 163 ; W. Carleton, Traits and Stories of the Irish Peasantry, ser. 2, Dublin, 1833, vol. iii, pp. 441-443. Armsrrone AnD LAwLor—The Domnach Airgid. 109 her it was purchased by Mr. George Sinith, the well-known Dublin book- seller. When it was opened it was found to contain a mutilated copy of the Gospels. It was apparently in Mr. Smith’s hands when it was exhibited to the Academy in 1832 ;? and a few years later it was sold by him to the Hon. Henry Robert Westenra, afterwards third Baron Rossmore, for £300. Ulti- mately it was purchased by public subscription, and became the property of the Academy in 1847. When the Domnach came to Dublin it was naturally assumed that it had originally belonged to St. Tigernach’s monastery at Clones, for nothing was then known of its early history except what might be inferred from the inscription which states that its outer case was constructed by permission of a comarb of Tigernach. That was the view expressed by Sir William Betham a day or two after he saw it for the first time. He wrote that the manuscript which it contained was “probably the property of St. Tigernach himself.” He seems to have found confirmation of this opinion in the representation on the cover of the shrine of one ecclesiastic handing a book to another. This, he held, was St. Tigernach passing on the Domnach to his successor Sinellus.* I mention the fact because it is an illustration of the fatal ease with which evidence for any theory can be found in such sources. But LBetham’s guess was put out of court by the discovery made by Dr. Petrie, that in two early documents—in one of which it was actually ealled the Domnach Airgid, or Silver Shrine—the Domnach was stated to have been presented by St. Patrick to St. Mac Cairthinn, the founder of Clogher. I quote the passages, as I shall have occasion to mention them in the sequel, The first is from the Tripartite Life of St. Patrick (ed. Stokes, p. 175£.): “Once as Patrick was coming to Clochar from the north, his champion, to wit, Bishop Mace Cairthinn, lifted him over a difficult place. This is what he said after lifting Patrick: ‘Oh, oh!’ ‘My God’s doom!’ saith Patrick, ‘it was not usual for thee to utter that word. ‘I am [now] an old man and I ‘ Tt is stated in the Transactions that Dr. Petrie read the paper, which will be referred to immediately, on 22 October, 1832, and 9 January, 1887. If so, it must have been largely altered in the interval between those dates. In the Ms. minutes, under 22 October, 1832, there is no mention of a paper: it is merely recorded that Petrie exhibited the shrine. But under 9 January, 1837, we read, ‘‘ Dr. Petrie exhibited a ms. of the Four Gospels of which he had given an account in a paper some time since before the Academy.” 2 Proceedings, iii, 237, 413 ; iv, 115. 3 Carleton, /.c., p.439f. That Sinellus was the successor of Tigernach, or had anything to do with the establishment at Clones, is unsupported by evidence. See my Fragments of the Register of Clogher in the Louth Archaeological Journal, vol. iv, no. 3, Extract ii. Sinellus was the possessor of a noted shrine, but it was called Deargann. 110 Proceedings of the Royal Irish Academy. am infirm,’ saith Bishop Mace Cairthinn, ‘and thou hast left my comrades in churches, and I am still on the road.’ ‘I will leave thee, then, in a church,’ saith Patrick, ‘that shall not be very near, lest there be familiarity (2), and shall not be very far, so that mutual visiting between us be continued.’ And Patrick then left Bishop Mace Cairthinn in Clochar, and with him the Domnach Airgit which had been sent to Patrick from heaven when he was at sea coming towards Ireland.” The second passage is from a Life of St. Macc Cairthinn, a portion of which is preserved in the Codex Salmanticensis (ed. C. de Smedt and J. de Backer, col. 799): “ Denique cum ipse in uniuersali praedicaret semen uitae per Hiberniam serendo iuuenilis annos aetatis suae transegisset, die quodam sanctum magistrum trans flumen quoddam ut solebat portauit, cumque sacrum deponeret onus signa debilitatis suspirando ostendit. Cum a pio patre pulsaretur ut causam suae diceret debilitatis, ait, Jam iniquit pater, accedente senio uires michi deficiunt et assiduum iter me multum grauat. Fae ergo me pater in aliquo loco Deo et tibi seruire. Quod pater annuens ait, Uade in pace fili et monasterium tibi construe in platea ante regalem sedem Ergallensium resurrecturus in gloriam inde. Sedes enim illorum qui terrena sapiunt deso- labitur, tua uero de die in diem augmentabitur, atque de eius sacro cimitherio plurimi ad beatam resurgent uitam. Et addidit, Accipe, inquit, baculum itineris mei quo ego membra mea sustento et scrinium in quo de sanctorum apostolorum reliquiis et de sancte Marie capillis et sancta cruce domini et sepulcro eius et aliis sanctis reliquiis continentur. Quibus dictis dimisit eum cum osculo pacis, paterna fultus benedictione. Itaque illue perueniens Clochorense fundauit monasterium.' Whatever else these passages may prove, they certainly show that when the documents from which they are taken were written the Domnach was not at Clones but at Clogher, and that it was believed by the writers to have been there from the foundation of the see. And the evidence of the Tripartite Life carries us back to a date long prior to the inscription to which reference has been made. The presumption therefore is that the shrine originally belonged to Clogher. ! The opening words of this extract are not in the Salamanca fragment, which begins at transegisset. They are taken from the Clogher Register. See below, p. 121. I print here part of the corresponding ge of Jocelin’s Vita S. Patricii, c. 143. It is no doubt derived from the Tripartite Life, and has no independent value. ‘ Expletis aliquantis diebus praefecit illum [Mac Cairthinn] Clochorensi ecclesiae quam ipse S. Patricius tune regebat, et cum consecrasset eum in episcopum, contulit illi chrisma-~ torium quod caelitus sibi destinatum susceperat.’ Here chrismatorium is probably a mere conjectural rendering of Domnach Airgit, ARMSTRONG AND LAwLOR—The Domnach Airgid. 111 Assuming for the moment that it remained there till the sixteenth century, I may indicate what seems to have been its later history. It was probably removed from Clogher in the interval between the death of Cornelius Mac Ardghail, who was alive and administering the see in 1592, and the arrival of the first bishop appointed by the Crown, the Scotsman George Montgomery. He succeeded MacArdghail, after a long vacancy, in 1605. From that time, except during its sojourn at Florence Court, till its purchase by Smith—a period of 130 or 140 years—it was in the territory of the Maguires, and in the custody of various members of the sept. Dr. Petrie’s discovery was apparently made shortly before he wrote, or finally revised, the paper in which for nearly sixty years he was generally held to have said the last word on the Domnach Airgid.* It was published in the Transactions of the Academy in 1838.° In that paper be found some difficulty in escaping from the thrall of the earlier theory. ‘hus he writes that the shrine was purchased “in the neighbourhood of Clones, in the county of Monaghan, tts original locality.” And further on he says that it ‘‘ belonged to the monastery of Clones, or see of Clogher”’; adding, in explanation of that obscure, if not unmeaning, phrase, that on “the death of St. Mac Cairthinn in the year 506, [St. Tigernach] removed the see of Clogher” to Clones’—a statement which is unhistorical. In this essay Dr. Petrie’s main purpose is to prove that the manuscript which was found in.the Domnach Airgid belonged to St. Patrick, and that it was brought to Ireland by him.’ His argument rests on the three following propositions :—1. he Domnach itself was the property of St. Patrick ; 2. It was originally intended to be a book-shrine, not, as has been commonly supposed, a reliquary; 3. The book for the preservation of which it was designed was the manuscript of the Gospels which was found in it in 1832. 1 Journal of Kilkenny Archaeological Society, N.S., vol. i (1856-7), p. 81. This conjec- ture, which was suggested by the historical facts, is confirmed by Mr. Armstrong’s opinion that the latest part of the outer case of the shrine dates from Elizabethan times, and that the figures with which it is adorned resemble the Maguire arms. There is no improbability in the hypothesis that the case was made after 1592, when the shrine was in the hands of the Maguire clan. 2 The evidence which connected the Domnach with St. Mae Cairthinn was unknown to Betham after the exhibition of the shrine in 1832, at which he was present. Carleton, l.c., p. 440. 3 Transactions of R.I.A., vol. xviii, Antiquities, pp. 14-24, +P. 14. On p. 17 Clogher is said to be the see to which it *‘ originally appertained.” The words ‘‘in the neighbourhood of Clones ” are misleading, for Brookeborough is almost equidistant from Clones and Clogher—about twelve miles from each, as the crow flies. The important point is that neither Clones nor Clogher is in the Maguire Country. 5P. 18. I may again refer to my Fragments of the Clogher Register, l.c. GIP, F1), See R.I.A. PROG., VOL, XXXIV, SECT. C. [18] 112 Proceedings of the Royal Irish Academy. Tf all these assumptions are sound, Dr. Petrie’s thesis is established ; if any one of them is false, it falls to the ground. All three were challenged in a valuable paper contributed to our Transactions in 1893 by Dr. J. H. Bernard, now President of the Academy and Archbishop of Dublin. Let us examine them one by one. 1. The first proposition is founded on the passages already quoted from the ‘l'ripartite Life of St. Patrick and the Life of St. MacCairthinn in the Codex Salmanticensis. And the statement which they agree in making, that St. Patrick gave the Domnach to St. MacCairthinn, must be admitted to be an early tradition. Nevertheless it cannot be true. Petrie himself tells us that the ornamentation of the inner metal case of the shrine—the Domnach Airgid proper—“ indicates a period between the sixth and twelfth centuries ” ;* and Mr. Armstrong now expresses the opinion, based on surer ground, that the seventh, or even the eighth, century is its earliest possible date. A shrine which was constructed in the sixth or any later century cannot have been in the hands of St. Patrick. But Petrie provides himself with a way of escape out of this difficulty. He gives us to understand that the wooden box within the Domnach is earlier than its most ancient parts, and he says plainly that it “may probably be coeval with the manuscript which,” according to him, “it was intended to preserve.”? Now it is obvious that the box was not designed for the purpose which it has actually fulfilled—to be the core of a metal shrine. For that end its sliding lid would have been useless ; once the box was encased it could not be drawn off. Moreover, if it lay on the rebate provided for it, its projecting end would have made it impossible to attach plates of metal to the side over the edge of which it protruded. These difficulties have not been satisfactorily surmounted by fixing the lid to the front of the shrine; for of necessity both it and one of the ends of the box fail to extend completely over the plates which they support, and the rebate which was intended for the lid has no function. It seems, therefore, that Petrie’s assumption that the box is older than the case which contains it is correct. But for the larger assumption, that it was already made in the fifth century, these facts give us no ground. Indeed, the hypothesis that this rude box, which appears never to have had a lid securely fastened to it, continued for three centuries, or even one century, to preserve a book, or relies, or anything else that belonged to St. Patrick, and was only after so long a period provided with a protecting cover, is in the highest degree improbable, But, assume its possibility, what becomes of the testimony of the Tripartite Life and the Life of St. Mac Cairthinn? One of ' Transactions, vol. xxx, pp. 303-312. 7 P, 15, 3 Thid, ArmsrronG and LawLtor—The Domnach Airgid. 113 them speaks of a shrine, the other of a silver shrine; and neither can be identified with a simple box of yew. If we try to re-habilitate the tradition by assuming that originally it applied to the box, and was later on transferred to the shrine in which the box was ultimately encased, we resort to conjecture. And mere conjecture, however plausible, cannot serve as an argument for Petrie’s purpose. The fact is, as I venture to think, that we have no warrant for the belief that St. Patrick ever saw any part of what we call the Domnach Airgid. But we have by no means done with a tradition when we have decided that it cannot be accepted at its face value. Every tradition of respectable antiquity will carry us some way towards the ascertainment of historic fact, if we subject it to a sufficiently searching analysis. At this point, therefore, we may attempt to ascertain whether the story on which Dr. Petrie laid somewhat undue stress throws light on the early history of the Domnach. Its main value, as I conceive, is that it fixes a later limit, for the date of the construction of the shrine. The tradition of the donation of the Domnach is undoubtedly early. The Tripartite Life, in which it first appears, is a work of the eleventh century, and “many if not all” of the documents on which it was based “ were composed before a.pD. 1000.” ‘hus the story, in the form in which the Tripartite Life presents it, had found its way into a written docu- ment by the tenth century. But the narrative in the Life of St. Mac Cairthinn, though agreeing in the main with that of the Tripartite Life, so far differs from it in detail as to suggest that the one was not derived from the other. The two accounts seem rather to be bifurcations of a tradition older than either of them.* Thus the Codex Salmanticensis makes mention of certain relics which were in the shrine, about which the Tripartite Life is absolutely silent. Doubtless one or more of the items in the list are insertions; but we have no reason to suppose that the general statement that the Domnach contained relics is relatively late. Again, the Life of St. Mac Cairthinn records that ‘For the date see Stokes, Vripartite Life, Int., p. lxiiff. Apparently the latest historical event referred to does not bring us further down than 936. This points to the early years of the eleventh century. 2 It may be pointed out that the argument is independent of the question whether the story of the donation is an intrusion in the Codex Salmanticensis. It is merely regarded as a recension of the story which differs from that of the Tripartite Life, quite apart from its date and context. But if it be a genuine portion of the text, what is here said of it may be applied to the whole narrative. The Tripartite Life implies that the Church of Clogher had already been founded by St. Patrick ; the Codex Salmanticensis states that it was founded by St. Mac Cairthinn. 3 This must be the case if Petrie’s remark (p. 20) is correct, that ‘‘some, at least, of the relies. . . . were not introduced into Ireland before the twelfth century.” He was no doubt thinking of the fragment of the Cross. Compare Mr. Armstrong's remarks, aboye, p. 104. es [18*] 114 Proceedings of the Royal Irish Academy. St. Patrick’s staff was given to Mac Cairthinn along with the shrine, while the Tripartite Life makes no allusion to the staff in this connexion. On the other hand, the Tripartite Life adds the statement that the Domnach “had been sent to Patrick from heaven when he was at sea coming towards Ireland.”? This of itself raises the suspicion that the staff had originally a place in the story; for the words are an accurate summary of what we are told elsewhere in the same work about the Bachal Isu. When Patrick had parted from Germanus, and was voyaging to Ireland, at Mount Hermon, in the neighbour- hood of an island on which he had stayed for three days, “the Lord appeared to him, and told him to go and preach to the Gael, and gave him the staff of Jesus.” Now at least as early as the twelfth century the staff of Jesus was one of the principal treasures of the Church of Armagh.’ It was therefore to be expected that a biographer of St. Patrick a couple of centuries earlier would suppress the statement that it was given to another see. Thus the origin of the story may be pushed back a considerable distance before the document used in the Tripartite Life—let us say to the middle of the tenth century. But again such a story, even if it were deliberately invented, could not gain currency within a few years of the making of the shrine. It cannot be supposed to have originated until the circumstances of the construction of the shrine and the name of its maker had passed into oblivion. The shrine can hardly be less than a century earlier than the tradition which made it coeval with St. Patrick and St. Mac Cairthinn. Hence we may conclude, with some confidence, that the latest date which can possibly be assigned to it is the latter half of the ninth century. But we may go further. A tradition such as we are considering cannot have come into being by spontaneous generation in a vacuum. Before it arose there must have been certain facts or beliefs which were the seed out of which it sprang. Before men could have come to believe that the Domnach was given to St. Mac Cairthinn they must have assumed that it once belonged to St. Mac Cairthinn, and that again implies the belief that it was then, and had been from time immemorial, at St. Mac Cairthinn’s church of Clogher. Thus it appears that the shrine had been at Clogher for a long period before the tradition took shape in its simplest and earliest guise. ‘The interval required for the evolution of the story of the donation out of the belief that the Domnach had belonged to St. Mac Cairthinn cannot, of course, be determined. We shall regard it as longer or shorter according as we suppose the tradition to have developed in the natural course, or the later story to have been a deliberate fabrication. Our view will also be affected by the opinion which we may hold as to the antiquity of the belief that ‘ Tripartite Life, p. 31. 2 sbid., p. 171. °St. Bernard, V. S. Malachiae, 24. Armsrrone AnD LawLor—The Domnach Airgid. 115 St. Mac Cairthinn was a disciple of St. Patrick. But in any case it cannot have been very brief; and the necessity of allowing for it carries back the date of the construction of the shrine to the confines of the eighth century, which Mr. Armstrong has already indicated as a terminus a quo. We seem, therefore, to have firm ground for the assertion that the Domnach was made in that century, and that, if not made at Clogher, it was deposited there shortly after its construction. In the church of Clogher it would seem to have remained from that day up to the eleventh or twelfth century. 2. We come now to Petrie’s second proposition, that the Domnach was originally a book-shrine. In favour of this hypothesis he adduces no other evidence than its form. “The form of the cumdach,” he says, “indicates that it was intended to receive a book.”! This is obviously disputable. Dr. Bernard was of the contrary opinion. ‘l’o him its form suggested, not a book-shrine, but a reliquary.? For myself I can only say that the examples of shrines undoubtedly made for the purpose of holding books do not seem sufficiently numerous to warrant a dogmatic statement on the ground of form alone. On the whole, I am inclined to agree with Dr. Bernard. The argument is at any rate not conclusive. But when we turn to evidence other than the dimensions of the case, we find that the only witness mentioned by Petrie tells against him. The Life of St. Mac Cairthinn, as we have seen, gives a list of the contents of the shrine: they are all relics; there is no suggestion that there was, or had been, a book within it. Petrie makes two attempts to get rid of this testimony. In the first place he asserts that the relics were not in the Domnach proper, but in recesses in the outer case. But that explanation of the words of the Life cannot be maintained. The writer tells us that the relics were in, not outside, the shrine. Moreover, the Codex Salmanticensis is a fourteenth-century manuscript. It is, therefore, quite possible that no part of the present outer case was in existence when it was written. But let us grant that it was already made. Then it must be noted that the enumeration of the relics is put into the mouth of St. Patrick. Assume everything that is in favour of Dr. Petrie’s contention: that the sentence before us was written towards the end of the fourteenth century; that it was not copied from an earlier exemplar—that of the body of the text or any other— but was actually composed by the scribe himself; that the scribe was aware that the case, with its attached reliquaries, had been already constructed: still the sentence cannot be dated much more than half a century after the making of the case, and the writer must have known that it was a A Te it). 2 P. 306. 116 Proceedings of the Royal Irish Academy. recent addition. How then could he assign it to the time of St. Patrick ? It is clear that he is speaking, not of it, but of the Domnach itself. Finally, only one of the recesses contains, or probably ever contained, relics. That one, it is true, holds a fragment of the Cross. But where were the relics of the Apostles, where was the Virgin’s hair (remembered as late as the nine- teenth century), where were the fragment of the Holy Sepulchre and the other relics, not definitely described ?! But Petrie has another defence. “Monkish biographers,” it seems, could not have known what the Domnach contained, for no one would have dared to open it. “No superstition was and is more common in connexion with the ancient cumdachs than the dread of their being opened.’* That may be. But that this cumdach was actually opened at the beginning of the century in which the Codex Salmanticensis was written is certain. The evidence will be produced lower down. The fact is that the documentary evidence is too strong for Dr. Petrie’s hypothesis. Viewed without prepossession, and in the light of the facts disclosed by Mr. Armstrong's investigation, it will be found, as I believe, to lead to the following conclusion. Prior to the middle of the fourteenth century the Domnach Airgid was used as a receptacle for relics, of which it contained a large store. When the outer case was made, the fragment of the Cross was removed from the interior of the shrine to a more fitting position, immediately above the representation of the crucifixion, which was the prin- cipal ornament of the new cover. There it was securely fixed in a recess, and covered by a crystal; and there it still remains. 3. The third proposition which is essential for Dr. Petrie’s argument is one of which he gives no proof, that the manuscript which was in the shrine a century ago was “the treasure for whose honour and preservation” the Domnach was made? Dr. Bernard makes short work of it. “It is demon- strable,” he writes, “that the inner case of yew was not made to contain the Gospels in their original form. For the measurements of the inside of this case are 7§ x 52 x 2} inches only; while the Ms., now that it has been opened, is seen to have been certainly not less than 9 inches long by 6} inches broad.”* When I read this more than twenty years ago, it appeared to me conclusive.° ! Dr. Petrie seems to have thought that the relics mentioned in the Codex Salmanti- censis were under the crystals on the outer case. The hypothesis was a priori probable. There was a relic under one of the crystals of the shrine of the Cathach of St. Columba (Transactions, xxxiii C, p. 394), as Petrie doubtless knew ; and below one of the crystals of the Domnach there is a fragment of the Cross. But of the six crystals on the Domnach three are on the plates which are of Elizabethan workmanship: the scribe of the Codex Salmanticensis cannot have alluded to them ; and whether there were any crystals in the portion of the shrine which they replaced we are ignorant. 220. eile: ‘P. 306 F. 5 Cp. Bury, Life of St. Patrick, p. 309. Armsrroné AND LAwLor—The Domnach Airgid. alee But lately a doubt has insinuated itself into my mind. It is supposable that the manuscript, when a shrine was made for it in the eighth century, had suffered considerable injury, and that its leaves were already frayed and eurled up. If so, measurements taken after the leaves were flattened out would not demonstrate that it was not the tenant for whose reception the box of yew was intended. And the very fact that a mutilated and unsightly Gospel book was deemed worthy of an elaborately ornamented cumdach would indicate a tradition that it had belonged, if not to St. Patrick, at least to some renowned saint. With this suspicion forcing itself upon me I was anxious that fresh evidence should be found. And a happy accident soon brought it into my hands. Last June I undertook to edit for the Louth Archaeological Society the extant fragments of a Register of Clogher which had been inspected by Archbishop Ussher in his early manhood, and about the same time by George Montgomery, Bishop of Clogher, but which has long since disappeared. The Register was compiled as late as 1525; but a study of the extracts from it which still exist convinced me that parts of it were based on—in one instanee copied from—a Register of Matthew Mac Cathasaigh, who was Bishop of Clogher for some years before and after 1500. Among the passages derived from this source is one which gives strong corroboration to the con- clusions which Dr. Bernard reached in his paper on the Domnach Airgid.! ' For the statements in this paragraph and those that follow about the Clogher Register, see my edition of the fragments in the Louth Archaeological Jowrnal, vol. iv, No. 3. The passage summarized in the next paragraph is Fragment iii, and runs thus :— Memorandum quod Dominus Matheus episcopus Clochorensis, init ad ecclesiam beati Constantis de Eoynys ad faciendam translacionem beati Constantis praespiteri et etiam beati fergiuminth episcopi de Cutmaine [/. Culmaine] successoris beati Aedhami Diaconi et invento corpore beati Constantis divisit relliquias eius in tres partes et relicta tertia parte in ecclesia de Hoynys duas secum duxit ad ecclesiam Clochorensem quarum vnam concessit ecclesiae beati Vumei de cuipsnech et fecit illam partem recondi in quodam scriniolo in quo similiter possuit episcopus de relliquiis supradicti fergiuminth episcopi Aliam autem partem de reliquiis Constantis fecit recondi in serinio magno beati Makartini in quadam cista parua siue pixide lignea. Item reliquias beati fergumintht} possuit in quodam linteo et fecit eas recondi in eodem scrinio. Item beatus Constans secum duxit de Britanneat tres viros praelatae fidei ad serviendum sibi et deo in monasterio suo de Hoynys quorum corpora sepulta sunt in inferiori parte Cymiterii de Hoynys et episcopus fecit vnum de ipsis transferri cuius relliquias possuit episcopus in quodam scrinio, quod dicitur Membra a Membrana dictum quae ibidem continebatur quia cum beatus Tigernacus hospitaretur apud lysdoniam vbi tune conventus monialium fuerat sanctarum qui conventus ynam habens ovem cuius lana induebantur moniales fecit interfici lam oyem beato Tygernaco pro carnibus et hoc ignorante sancto Tygernaco: wnde beatus Tigernacus rogauit Dominum suum Jesu Christum vt pellis illa afferet lanam ad ysum monialium sicut antea dabat, quando fuit in corpore ovis, Et sic conyentus habebat lanam de pelle per plures annos sicut viuente oue ad supplicationem beati Tygernaci. Postea vero cum pellis illa ad solem fuisset possita causa eam siccandi coruus quidam pellem in rostro recipiens eam 118 Proceedings of the Royal Irish Academy. Tt tells us that Bishop MacCathasaigh went to the island of Eoinis for the purpose of translating thence the relics of St. Constans and St. Fergiu- minth of Culmaine. The relics of these saints were divided into three parts, one part being brought to Clogher. The bishop at the same time disinterred the relics of a follower of St. Constans, and carried them also to Clogher. The relics of St. Constans were there deposited “in the great shrine of St. Mac Cairthinn in a certain little box (cista) or pix made of wood.” The relics of St. Fergiuminth were placed in the same shrine wrapped in linen (in quodam linteo); while those of the unnamed follower of St. Constans were put into another shrine known as (quod dicitur') the Membra. We are given our choice of two derivations of that obscure word, one of which is interesting as showing that the shrine was large enough to contain a sheep-skin. Accord- ing to the other derivation Membra is equivalent to memoriale scrinium ; and this huge shrine was so named “in respect of the greater shrine which is called (dicitur!) Domnach Airgid. In which memorial shrine relics of saints were deposited.” Then follows a sentence which may perhaps be rendered, “But that great shrine was intended (mittebatur') for arduous matters, but this for minor matters.” The writer seems to mean that the Domnach Airgid was called great, not because of its size—for it was evidently the smaller of the two—but because of the veneration in which it was held: it was the repository of the more important relics, the Membra of the less important. The passage concludes with the statement that the anniversary of the translation, which took place on 6th September, 1308, was ordered by the bishop to be observed as a festival. This appears to be the only known document, except the Tripartite Life, in which our shrine is called Domnach Airgid; and it is the only one which secum per aera reduxit vsque ad sanctum Biedanumt et dimissit pellem ad pedes eius, qui sanctus Biedanus+ cognoscienst pellem esse de qua fecit sanctus Tygernacus miracula remissit eandem pellem ad ecclesiam Clochorensem Cui tunc praefuit beatus Tygernacus episcopus et sic scrinium illud factum est ad opus illius pellis in quo pellis illa recondita fuit. vel alio modo dicitur Membra id est memoriale scrineum respectu maioris scrinei, quod dicitur Dornnach a:ipgero. in quo scrineo memoriali reliquiae sanctorum fuerunt reconditae vnum [/. verum /] illud magnum scrineum ad ardua negotia mittebatur. Hoc autem ad minora negotia. Translacio quidem supradicta fuctat fuit apud Eoynys octavo Idus septembris Anno Domini 1308 quam diem praecipit episcopus obseruari pro factot [l. festo] translacionis plurimorum confessorum. 1 Note the present tense. Both shrines would seem to have been in existence and still at Clogher in 1525. 21 am doubtful about this translation. Perhaps mittebatur means ‘was sent,’ i.e. from Clogher to outlying places, for the taking of oaths, especially of expurgation. The Domnach was used in this way in the last century (Mason, I.c.). But the whole context shows that the more important relics were reserved for the Domnach, and that would be sufficient reason for ‘sending’ it only on more difficult or remarkable occasions. ARMSTRONG AND LAWLOR—TZhe Domnach Airgid. 119 gives contemporary witness to an incident in its history prior to the nine- teenth century. It is of great interest. It proves that a few years, at the most forty-five, before the outer case was made the Domnach was not at Clones but at Clogher, and it implies that Clogher was its normal home. We may infer, with little hesitation, that it was there from the time of the writing of the Tripartite Life of St. Patrick and the Life of St. Mac Cairthinn to the year 1308. Moreover in that year it was the principal shrine of the cathedral church. It is most unlikely that the bishops surrendered it to the Abbot of Clones shortly afterwards. Indeed the inference is probable that it was retained in the cathedral as long as the see was held by bishops who set a high value on a shrine so ancient, and on the relics which it contained, that is to the last decade of the sixteenth century. The document itself seems to indicate that it was still in its old home in 1525, when the Clogher Register was compiled. Why the permission of the comarb of Tigernach should have been required for the construction of the new case it is impossible to determine; but the inscription which records this fact can no longer be held to prove that it belonged to Clones.’ Further, the alternative title of the Domnach was the “great shrine of St. Mac Cairthinn.” This makes it fairly probable that the shrine was supposed to be as old as the fifth or early sixth century, and it confirms the opinion already expressed that at the beginning of the fourteenth century it was believed to have been always the property of the church of Clogher. But in view of the inscription just mentioned, the omission of any reference to St. Tigernach is significant. Hardly less significant is the silence about St. Patrick. We may venture to infer from it that the story that the Domnach was a gift from St. Patrick to St. Mac Cairthinn was unknown or disbelieved at Clogher. Again, the shrine was obviously used as a reliquary. The relies of St. Constans and those of St. Fergiuminth were laid in it, the former enclosed in a “cista parua siue pixis,’ the latter wrapped in linen. There can be no doubt of the fact if the “ cista parua” is the existing box of yew; but this is improbable.? It is certain, however, that neither the small cista nor the linen wrapping can have been attached to the outside of the case. Nor can they have been placed beneath crystals, even if we suppose that the original metal covering had such adornments. Be it remembered that here we have to do 1 See further below, p. 126. 2 For these reasons: (1) it would hardly have been called a *‘parua cista” ; (2) the context suggests that the ‘‘cista” was made for the reception of the relics: the yew box was in the shrine from the first ; (3) the relics of St. Fergiuminth were apparently not in the ‘‘cista’’; but they could not have been in the shrine and yet outside its lining. R.I.A. PROC., VOL. XXXIV, SECT. C. [19] 120 Proceedings of the Royal Irish Academy. with the Domnach Airgid proper: the outer case had not yet been made. And the statement of the document is express: ‘aliam partem de reliquiis fecit recondiin serinio.’ Further, there is no hint that there was a book in the receptacle into which the relics were put. And there is nothing in the memorandum to suggest that in 1525 the shrine was used for any other purpose than that which it fulfilled in 1308. Finally, there can be no question, in spite of Petrie’s assertion that such an act was impossible, that on this occasion the Domnach was opened, and that the eyes of at least a few favoured ecclesiastics saw what it contained. There is no improbability in the assumption that “monkish writers” shared the knowledge which they obtained. We may now turn to another part of the Clogher Register, in which we might reasonably have expected to find mention of the Domnach. In the course of this paper I have frequently referred to a fragment of a Life of St. Mac Cairthinn in the Codex Salmanticensis. Now in 1528, three years after the main part of the Register had been completed, Patrick O’Cuillean, Bishop of Clogher, added to it an office for the festival of the patron saint of the diocese. There remain of it the hymn and the lesson.' The latter is a short Life of St. Mae Cairthinn, which has an evident relation to the Salamanca fragment. The differences between them are not such as to forbid the belief that they are two recensions of the same original. Apart from blunders of the scribes they fall under three heads. In the Register there are many slight variations from the Codex Salmanticensis, which may be described as stylistic, alterations of the order of the words, and so forth, which do not alter the sense. Secondly, the Register omits four passages and abridges others, all of which are found in the Salamanca copy: of the omitted pieces one was certainly in the ultimate, if not in the immediate, exemplar of the lesson in the Register. And lastly, the Register adds a paragraph at the end, . relating the appearance of the saint in a vision to a woman, and mentioning his death on 24th March. We have sufficient reason to believe that these variants were changes deliberately made by Bishop O'Cuillean in the text of the lesson ; for at the end of the Office he wrote the following note :— “ Suprascriptum officium fuit hie per Reverendum uirum Patricium Culinn episcopum Clochorensem ad utilitatem publicam ecclesie Clochorensis redactum ex antiquis libris ordinis sancti Augustini studiose transcriptum, et secundum usum Romanum in melius ordinatum atque reformatum superflua reticendo ac diminuta augendo et perfectius emendando, Anno Domini 1528.” Here the bishop avows that he altered the text, and he divides his changes under the three heads of omissions, additions, and emendations, no doubt of style. } Extract xi in my edition. Armsrrone anp Lawtor—Vhe Domnach Airgid. 121 Now one of his omissions is very remarkable. It is in the passage relating to the foundation of the monastery of Clogher. I print it in full in order that the reader may be able to compare it with the corresponding part of the fragment in the Codex Salmanticensis quoted above, p. 110. “ Denique cum ipse in vniuersali praedicaret sem[en] vitae per hiberniam serendo juuenilis annos aetatis suae transegit die quadam Patricium patrem more solito transiens per quoddam flumen et portans sacrum onus deponendo suspirauit et percontatus a pio patre vt quid suspiraret inquit pater accedente senio vires meae deficiunt, et assiduum me grauat iter, fac igitur in} si placet in vno loco deo et tibi servire, sanctus Patricius ait, placet ait vade in bonet et monasterium construe in platea ante regalem sedem Ergallencium inde resurrecturus in gloria, illa desolabitur, tua vero sedes de die in diem augumentabiturt de cuius sacro Cymiterio plures ad beatam Vitam sunt resurrecturi. Deinde vir sanctus ad praefatam plateam pervenieus Clochorense fundauit monasterium.” In this extract there are many minor departures from the Salamancan text. Dut far more important is the complete excision of two consecutive sentences, the first of which begins, “ Et addidit, Accipe, inquit, baculum.” Bishop O’Cuillean has actually omitted all reference to the gift by St. Patrick to St. Mac Cairthinn of his staffand shrine. How can this be accounted for? Most readily, I think, on the supposition that the bishop rejected the old legend of the donation of the Domnach. If he had believed it to be a true story, he would hardly have omitted it. It could not have been regarded as one of the superfluous things which, as he tells us, he passed over in silence.’ It will be remembered that we have already had reason to doubt that this tradition was accepted at Clogher in 1308; and the story told 600 years later, that the Domnach was brought from Rome by Donagh O’Hanlon, is proof that it was not accepted in the Maguire country in the nineteenth century. It is worthy of note, in this connexion, that in none of the extant fragments of the Clogher Register is St. Patrick claimed as founder or first bishop of the see, though St. Mac Cairthinn is said to have been his disciple and “fortis athleta.” In the Catalogue of bishops St. Mochta of Louth is the first,? and St. Mac Cairthinn the second, St. Patrick receiving only incidental 1 It is of course possible that this episode was not omitted by Bishop O’Cuillean, but inserted from another source in the Codex Salmanticensis. If that be the true account of the matter, we still have evidence that the Life of St. Mac Cairthinn used at Clogher knew nothing of the donation. * Extract ii. The name is latinized ‘‘ Matheus.’’ According to Ware the Register described Mac Cairthinn as the first bishop (De Praes. Hib., p. 41), but this is an error. Jocelin ignores St. Mochta, and makes St. Patrick the first bishop. See above, p. 110, note. [19*} 122 Proceedings of the Royal Irish Academy. mention. This is the more remarkable, inasmuch as St. Patrick’s preaching at Clogher is referred to, and the see-lands are regarded as in some sense belonging to him. Thus there is some reason to think that the tradition, ancient as it seems to have been, that the Domnach Airgid belonged to St. Patrick, was never current at Clogher. If so, its value for historical purposes is seriously diminished. It may well have been invented to support the ever-growing pretensions of the see of Armagh. I conclude by summing up the main results to which this investigation seems to have led us. It appears probable that the Domnach Airgid was preserved at Clogher from the time of its construction to the end of the sixteenth century. There is no evidence that it was a book-shrine. At the end of the thirteenth century it was certainly used as a reliquary. It is unlikely that at that period it held a book. The manuscript which was found in it in 1832 was probably put into it in some later century, just as some leaves of a manuscript which do not belong to it were placed in the eumdach of the Book of Mulling* Lastly, there is the gravest reason to doubt the historicity of the tradition which credits St. Patrick with having given the Domnach to St. Mac Cairthinn. That tradition cannot be invoked as a proof of the antiquity of the Gospel-book which it enshrined in later times. The question of its date must be settled, if settlement is possible, by the palaeographers. NOTE ADDED IN PREsS. When the foregoing paper was read before the Academy, my friend Professor R. A. Stewart Macalister made some striking remarks upon it, the substance of which he has kindly reduced to writing in the following com- munication :— I.—The following facts have been brought forward :— 1. There was in mediaeval times a shrine at Clogher known as Domnach Airgid, known to contain certain relics, including a lock of hair of the Blessed Virgin. One tradition, not apparently universally believed, ascribed its origin to St. Patrick. 2. This shrine was opened in the year 1308, and further relics (not a bovk) were added to those which it originally contained. The nature of these relics is specified at least in general terms in the Register of Clogher. 3. The shrine in the Academy’s collection, identified by Petrie with the Clogher shrine, was traditionally called Domnach, not Domnach Airgid, teste Petrie: it contained very sacred relics, nature not specified, feste the Statistical Survey. O’Beirne’s letter, appended to Carleton’s Traits and Steries, mentions a lock of hair of the Virgin as among its contents. It is curious that Petrie 1 Extract iv. 2 See my Chapters on the Book of Mulling, p. 12 f. ArmstronG AND LawLor—TZhe Domnach Airgid. 123 makes no reference to this tradition, which is the strongest argument in favour of his theory of the identification of the two shrines. 4, The inscription on which Petrie relied for the identification of the Academy shrine with the shrine of Clogher is now shown to have been mis- read; and the only historical inscription on the shrine connects it with Clones, not with Clogher. 5. The Academy shrine was found to have contents quite different from those known to have been in the Clogher shrine. II.—From these facts I draw the following conclusions :— 1. The Clogher shrine is not identical with the Academy shrine. The Clogher shrine, with its relics, is lost ; probably it disappeared at the Reformation. 2. The Academy shrine is a Clones shrine, to which no authentic tradition attaches. But the nature of the more important relics in the Clogher shrine was well known: and after the disappearance of the Clogher shrine the popular traditions with regard to them became attached to the Clones shrine and its then unknown contents. II1.—With regard to the contents of the Clones shrine, it is to be noticed— 1. That the book was a crushed, illegible fragment, not a carefully pre- served MS. such as we usually find in book-shrines (e.g. the Cathach, or the Stowe Missal). Therefore, either the book was itself preserved, not for study but as a relic; or else was a worthless bit of padding squeezed into the box when its legitimate contents were either lost or abstracted. Of the two theories, the former is by far the more probable. It explains at once what would otherwise be unintelligible—why were these fragments, from a literary point of view useless, deposited in the case ? 2. This being assumed, the following further deductions appear probable : a. That the book belonged, or was supposed to have belonged, to some important saint connected with Clones, and was there preserved as a relic of him. Most likely this would be Tigernach himself, the founder of Clones. b. That it was for a long time used as a wonder-working relic, and maltreated in various ways as such—cut up for amulets, soaked in water to be subsequently administered for curative purposes, &c. Such treatment would be sufficient to account for its condition. e. That after it had suffered serious injury from a continuance of this treatment, it was encased in the yew box, perhaps for convenience of trans- port rather than for preservation. This box could be opened, and the maltreatment of the Ms. continued as before. d. That in time the ecclesiastics of Clones realized that the relic would disappear under this treatment altogether, and therefore they hermetically sealed it up by enclosing the box in bronze plates. 124 Proceedings of the Royal Irish Academy. e. That, long afterwards, the silver outer case was added. It was not till this was done that the shrine could be called Domnach Airgid; it follows that it cannot be the shrine referred to under that name in the Vita Tripartita Patricii, which is a document much older than the outer case of the reliquary. J. That in time (probably after the Reformation) the true nature of the contents of the Clones shrine became forgotten, and it was then popularly eredited with being the receptacle of the far more important and impressive relics which the lost Clogher shrine had contained. The main point in this very interesting criticism is Professor Macalister’s contention that Petrie wrongly identified the shrine found at Brookeborough in 1832 (the Academy shrine) with the Domnach Airgid mentioned in the Tripartite Life (the Clogher shrine). It gives me the opportunity of supply- ing a defect in my paper, in which I have adopted Petrie’s assumption without argument. The evidence may be stated as follows :— 1. The word “ Domnach” is very rarely used as synonymous with “shrine.” It was applied to the Academy and Clogher shrines, and, so far as I know, to no others. 2. The epithet airgid, which was used of the Clogher shrine, suits the inner metal case of the Academy shrine: it was in appearance a silver box. It is of course true that this shrine had in strictness no title to the epithet before the silver-gilt plates had been added. But it must not be assumed that the name Domnach Airgid was devised by the maker of the shrine to which it was applied, or by others who knew the details of its structure. It is much more likely to have been a popular designation of later days. That ordinary folk should have supposed that the Academy shrine was made of silver is not surprising. Till Mr. Armstrong told us that its coating was of tin, all inves- tigators held that it was of silver. To be sure, they recognized that it was only a coating: but then they saw the bronze plates which it covered. Before the coating was worn off, and while the inside of the case was not in view, the shrine might quite naturally be dubbed “of silver.” But when the gilded covering was made, that description became obsolete. If after that date it had received its name, it might have been called Domnach dir. On Petrie’s hypothesis it is easy to see why in later centuries the epithet airgid was dropped, and it came to be known simply as ‘ the Domnach.’ 3. We have little evidence about the material of which shrines were made in early times. Shrines are frequently mentioned in the Annals; but apart from those of St. Conlaed and St. Ronan, “of gold and silver,” con- structed about 800, and the golden cumdach of the Book of Kells, stolen in 1006, on account of which the book “was the principal relic of the western ArmsrronG AND LawLor—The Domnach Airgid. 125 world,”! 1 can recall no instance in which the material of a shrine is specified. We may infer that these were exceptional shrines, and that the use of the precious metals, or what passed for such, was unusual. This conclusion is not disturbed by the legend that St. Columba desired a church “full of gold and silver to cover relics and shrines withal,”? We may believe, therefore, that “silver shrines” were rare, though we have an example, perhaps older than the Clogher shrine, in the Lough Erne shrine, which, like it, is coated with tin. The very name Domnach Carleton, Traits and Stories, 1853, vol. iii, p. 442. ~ 126 Proceedings of the Royal Irish Academy. King of Dartraige (in which is Clones), and other nobles, nearly all from the present county of Monaghan. The third was headed by Donn Maguire, King of Lough Erne.’ This grouping must represent the natural division of the diocese at the time. The close connexion which subsisted between Clogher and the western part of the diocese is indicated by the names of the principal members of the Cathedral Chapter. Thus, between 1450 and 1550 we know the surnames of five Deans—three Mac Cawells (a Tyrone name) and two Maguires. “Dean McCrenyr” (Shirley's Monaghan, p. 321) was apparently a native of Co. Monaghan, but it is not certain that he was Dean of the Cathedral Church. In the same period three Archdeacons are known— two Maguires and an O’Cassidy. O’Cassidy is a Fermanagh name, and in the fifteenth and sixteenth centuries the family was intimately associated with the Maguires (see Mac Carthy, Annals of Ulster, index). These facts seem to me to constitute a strong argument that the Academy and Clogher shrines are identical. The only difficulty in the way is the inscription to which Professor Macalister refers. But it does not prove that the outer case was made for a shrine preserved at Clones. If the shrine had been in the custody of the comarb of Tigernach, the case would have been made by his order, not merely by his permission. The shrine of the cathach of St. Columba—a Tirconnell relic—was made “by” Cathbarr O’Donnell and “by” Domnall MacRobartaigh, Comarb of Kells, and the artist who con- structed it lived at Kells The meaning of this seems to be that O'Donnell, by the good offices of the comarb, had the shrine made by a Kells worker. Our inscription may indicate no more than that, by permission of the Abbot of Clones, a metal-worker of his monastery executed the work which the authorities of Clogher desired. Whether he did the work at Clones or at Clogher is a question of no importance. It is unnecessary to criticize Professor Macalister’s arguments in further detail. But his final “deductions” must not be passed over in silence. They postulate for the manuscript found in the shrine a date long prior to the box of yew, which is itself assumed to be more ancient than the earliest metal plates. ‘The manuscript, on this theory, cannot be later than c. 650. It is suggested that it may have belonged to St. Tigernach, who died in 950. Palaeographers must decide whether the script is consistent with so early a date. But the text is Vulgate with Old Latin mixture.’ That such a text should have been in use in Ireland by 550 or even 650 may be pronounced at least improbable in view of the history of the Latin Scriptures in this H. J. L. country.‘ "2 Proceedings R.I-A., xxxiii C, 391. ' Clogher Register, Extract viii. 2 Transactions R.I.A., xxx, 310 ff. ‘ Proceedings R.I.A., xxxiii C, 314 f. If Professor Macalister’s date (cent. ix) for the Book of Durrow be accepted, the improbability becomes greater. PLATE ILI, XXXIV, SEcT. C. Proc, R. I. ACAv., VOL. ). Top (UPPER PLATE FRONT, ARMSTRONG AND LAWLOR.—THE DOMNACH AIRGID, ax : “jes71 Ce rar Proc. R. I. AcAp., VoL. XXXIV, Sect. C. PLATE IV. DEXTER SIDE. SINISTER SIDE. ARMSTRONG AND LAWLOR.—THE DOMNACH AIRGID. Proc. R. I. ACAD., VOL. XXXIV, Srcr. C. PLATE V. INTERIOR. BASE. ARMSTRONG AND LAWLOR.—THE DOMNACH AIRGID. [pena zara) WATT: THE EARTHWORKS, TRADITIONS, AND THE GODS OF SOUTH- EASTERN CO. LIMERICK, ESPECIALLY FROM KNOCKLONG J'O TEMAIR ERANN. 3y THOMAS JOHNSON WESTROPP, M.A. Read Fenruary 25. Published Aveusr 29, 1918. PLate VI. CONTENTS. PAG PAGE (A)—Tue BordeRLAND BEFORE ‘a (&)-—Tre Trina Gops: Bodbh History, . 0 F 29 Dearg ; Cairbre Musc ; Cliu ; Cuil; Nechtan ; Deda, Gar- (B)—Tue Gops ry Criv, 185 ban, and Febra; Dergthene ; : Donn ; Edaein; Ferfi and the Knockainey legends, . 151 (C)—Lrrerary TRACES OF THE Gops, 138 (F)—Tur Eartaworks: Knocklong, Aghadoon, Atheneasy, Bal- (D)—Locat Cunrus: Lugh ; Nuada linseaula, Temair Erann, . 170 and Manannan, . . q alent AppenpIx—Temair Luachra,. 181 THE subject of Irish local mythology’ has never, up to the present time, been studied on scientific lines. Save the great sanctuaries,* no attempt has been made to localize the centres of the worship of the Celtic gods. In Ireland, the usual method has been to bring together a crude heap of legends of all periods and districts, and try to bring order out of the chaos as it exists. Naturally no consistent results emerged from so unscientific a process; there was always a large waste product, as in an abortive chemical experiment, and 1T use the following abbreviations among more obvious ones:—R. Soc. Antt. Tr. = Royal Society of Antiquaries of Ireland, under its successive names ; Dind S. = Dind Senchas (Rennes and Metr., metrical) ; Hib. Lect. iv = ‘‘ Celtic Heathendom,” J. Rhys, Hibbert Lectures, iv, 1886 ; Ir. Myth. Cyc. = ‘‘ Irish Mythological Cycle,” H. D’Arbois de Jubainville (tr. Ric. Best, 1903); Encye. Relig., Professor MacCulloch, on Celtic Religion, Encyclopaedia of Religion and Ethics, vol. iii; ‘‘ Leab. Gabh” = Leabhar Gabhala i, MacNeill and Macalister; ‘“‘Silva Gad.” = ‘Silva Gadelica” (tr. S. H. O'Grady); ‘‘ Keating Hist.” = ‘‘ History of Ireland” (Ir. Texts edition); ‘‘ Agall” = Agallamh na Senorach (version in Silva Gadelica) ii. ‘‘ Holder,” Alfred, ‘* Alt- celtischer Sprachschatz.” 2 Since this paper was written, Professor Macalister’s important monograph on Tara has been read. R.I.A. PROC,, VOL, XXXIV, SECT. C. [20] 128 Proceedings of the Royal Irish Academy. this was ignored, or “ written off,’ with a light heart, as “inconsistent with the facts,” i.e. with some one group of facts. The fault is that such workers failed to see that we have stories told of the gods in various tribes (contra- dictory even where ancient), and a huge mass of corrupted theories and chronologies, later than A.D. 1050, formed without realizing its contradictory nature. There is no “in the beginning” in Irish mythology; and even of the ending, “the last great battle of the gods,”! we have a duplicated and irre- concilable account, not a clear-cut tale, like that of “the days when Ymir dwelt,” or of Ragnarék, in the Norse mythology. It is our duty to begin research, even if we have to penetrate many a labyrinth leading to nowhere, and to lament the loss of many a plausible system. A false theory negatived is a positive result. I make no claim for my own theories, but that they seem to contain the facts known to us at present; in no confidence, but in - much uncertainty, | am entering on an unexplored country. I was dis- appointed that my ‘Aine papers? (a portion of the present larger study, then nearly complete) drew me no helpful criticism. J only got evidence of the existence of a deep prejudice against folk-lore and mythology, in such criticisms as “ buttressing superstition” and “bringing legends into the study of forts.” Now we, alone in Western Europe, have a large mass of very early literary material to help us, and (anywhere but here) it is a truism that every ray of light should be concentrated on a subject. Ireland being “the last resort of lost causes,’ preserved record of a European “culture” as primitive as that of the South Seas, and therefore invaluable for the history of human advance ;? elsewhere its existence is only to be established from hints and equivocal survivals. Our early tales are no artificial fiction , but fragmentary beliefs of the pagan period, equally valuable for topography 'The ‘First Battle’ (Revue Celtique, xii, p. 57) mentions Norse coins, and is evidently ‘‘rationalized"’ and localized. Rhys hesitates to accept the battles as one, but the ‘‘ First” battle seems a euhemerized version of the other. De Jubainville (Ir. Myth. Cycle, pp. 84-9) argues for their identity. The two battles first appear with Flann of Monaster, cirea 1050. The battles coincide with the feasts of Midsummer and Samhain. Sanas Chormaic (ed. W. Stokes, p. 123) mentions one battle ; so does Cinaed Ua Articain (970), this being the so-called ‘‘ second” fight. See also Eriu, viii, pp. 17, 25, 45, and Squire, ‘‘ Mythology of the British Isles,” p. 75. 2 Supra, xxxiii, p. 9, p. 444; xxxiv, p. 47. 3 “ Survivals of Totemism among the Celts ” (Rev. Celt., xxi, p. 269) ; ‘‘ Human Head Trophies,” Celtic Rev. iii, pp. 68, 81. Encyc. Relig., vi, pp. 534-8. Rev. Celt., xxxiv, S. Reinach, pp. 38, 253. Human Sacrifice,’’ Eriu, ii, p. 86. Dr. Joyce, ‘‘Social History of Ancient Ireland,” i, p. 239. Human skulls were found under an altar inscribed to the Celtic war god, Encyc. Relig., vi, p. 534. Such heads are the ‘* mast crop of Macha.” “Three Irish Glossaries,’ W. Stokes, p.xxxv; Echtra Airt in Book of Fermoy, f. 139 ; Eriu, iii, p. 155. Ordeals, Irische Texte, 3rd ser., Ist pt.; tabus, notably the Book of Rights, the Death of Cuchullin, Bruden Da Derga; also paper by Miss E. Hull, Folk Lore, xii, No. 1. Westrropp—Zhe Hurthworks, §¢., of S. £. Co. Limerick. 129 and for mythology. Topographical comment—I will not say criticism—has been equally inefficient.’ A theory is not refuted by saying “ all the great antiquarians are against you,” “the Psalter of Tara (!) refutes that,’ or “Q’Donovan has set the question past all doubt.” These remarks only prove that we have hardly commenced scientific archaeology in this country. (A)—Tae BorDERLAND BEFORE Hisrory. Ireland being cut offso much from the world-centre in the beginning of our era, naturally got little record from Greece, Italy, or from those who have left us so little independent material about themselves as Gaul and Britain have done. A great geographer, about A.D. 160, collected the names and positions of several nations and places in our island. Beyond that, we only get a tale of an expelled chief, who, late in the first century, sought aid from a foreigner against his countrymen; a hint of an unrecorded landing “ beyond the shores of Juverna”; sketches of “the Scot” (doubtless our Niall of the Nine Hostages) “making the sea foam with hostile oars”; of his defeat and the erief of “icy Ierne” at the slaughter of her sons—-little else. Our native material in general was handed down (much, presumably, in mnemonic verses) by schools of bards and druids till Ireland, fully Christianized, reduced her tales to writing in the late sixth or earliest seventh century. There is little doubt as to our having some slightly altered material preceding, perhaps, the fourth century. Unfortunately this has been recast, so that even its broad outlines may be questioned. Ptolemy knows nothing of Tara or the Milesians ;* all his tribes are of the “tributary” or “non-Milesian” races® and the Ulidians and Picts. However, much of value escaped, for Roman culture never intervened, and Christianity in Iveland was a tactful, warm- hearted mother, and learned the stories to tell to her children. Of course, we should not presuppose that the ancient Irish had no writings. Their Gaulish and British kindred had them, Probably rituals and mysteries were not written ;' but genealogies, lists of chiefs, and even 1] discuss the one valid objection made to the identity of ‘Oenach Culi at the end of this paper. * This may arise from his only giving coast tribes. 3 T use this term ‘‘ Milesian” for the fair races presumably British, but commit myself to no theory in using these conventional names. Mil seems to have been a god, not unknown in Europe, so also was his father Bile. Rhys regards Mil as ‘ the Celtic Dis” (Hib. Lect., iv, p.582). Nuada was son of Beli in Welsh story. Beli was adored at Uisnech. *T have not noted any definite pagan ritual save the sunward turn, certain formal curses, the offerings of men, cattle, pigs, milk, corn, and brushwood, and the excessive prostrations before Crom Cruach [but, since this was written, Professor Macalister’s very suggestive paper on Tara gives much food for thought.] See preface to ‘* Three Irish Glossaries,” pp. xxxv-vi. [20*] 150 Proceedings of the Royal Irish Academy. tales may have been recorded briefly and on perishable materials ; the ogham was not suited for literature. Ethicus of Istria, early in the fourth century, claims to have examined books of the “ unskilled philosophers” and “ unculti- vated teachers” of Ireland. St. Patrick reckons “Gaulish rhetoricians ” among his opponents in Ireland; Juvenal tells us, “ Gallia causidicos docuit facunda Britannos.” Perhaps, these “spread the light” to Ireland. The “Yellow Book of Lecan” has a story how the saint burned 180 druidical books: and MacTheni (ante A.D. 698) tells of an ordeal by which Patrick and a druid tested the merits of their respective books.! To argue that the tales of the long survival of Fintan and Oisin imply a belief that there was no early written record is to go too far. Only sucha being as Fintan could tell of colonies lost in the Deluge or exterminated, to the last life, by a pestilence (indeed, the Fintan story is evidently a clerical recension of an archaic tale of rebirths); while the Oisin cycle is confessedly late, too late to give a clue as to belief in the early fifth century—the omniscient “ eagle (or hawk) of Achill” were as good an “ authority ”! Taking the less popular (and on that account less corrupted and more archaic) cycle of tales of the Red Branch, we have probably Ulidian legends of the opening centuries of our era, with no trace of the tribal pre- dominance of the Milesians of Tara, or the provincial arrangements, or the ancient political importance of that place, or the dawn of the Milesian tribes of Magh Femen, Cashel, and Thomond. The Ernai and Mairtene loom large ; the manners and customs are identical with those of the Gauls just before our era ;? and some of the legends® are outrages on all the ethics and prejudices of the clerical writers, and evidently remotely ancient and almost unexpurgated. Now, for our own district we have a great mass of tales from several different, and even hostile, tribes; no mere fictions, but once told to those who believed that gods dwelt on the great blue peaks of Galteemore and ** Cosmographia Ethici,"’ ‘‘ Social History of Ancient Ireland” (Dr. P. W. Joyce), i, pp- 403-5 ; cf. Juvenal, “‘ Satires,” xv; Petrie, ‘‘ Tara Hill” (Trans. R. 1. Acad., xviii, p. 47); ‘* Yellow Book of Lecan” (Celtic Review, v, p. 3) ; ‘‘ Tripartite Life of St. Patrick’’ (ed. Whitley Stokes, p. 44), from “‘ Book of Armagh.’ See also ‘‘ Battle of Magh Leana,” p. 21; ‘* Manuscript Materials (Eugene O’Curry), p. 505. The opening “ geasa” poem in the Book of Rights is purely pagan, save a ‘‘redeeming’ verse (pp. 1-23). See also Book of Ballymote, p. 190b). J. MacNeill (supra, xxvii, p. 333) gives evidence for early Christian hostility against Irish pagan culture. = **Celt. Rey.,” iii, p. 65 (Miss Eleanor Hull); ‘‘ New Ireland Review,” xxvi, p. 130, p. 84 (J. MacNeill) ; ‘‘ Duanaire Finn” (Ir. Texts Soc.); introd. (same). * Those of Duben, Clothra, Bresal, of Cnogba, and Ness; see R. Soc. Antt. Ir., xl, p. 184; Keating’s “ History,” vol. ii, p. 233; ‘‘ Metr. Dind S.,” x, p. 45, and ‘Coir Anmann ” (Irische Texte Ser. iii), p. 403. Wrsrropp—Zhe Larthworks, §c., of S. E. Co. Limerick. 131 Mish, on the hills of Luachair and Slievereagh, and even on low ridges like Knockainey and Knockfirina. The hearers gloried in a physical sonhood from Lugh or Nuada, The tales were probably largely preserved by a notable guild of Druids round Cenn Febrat, or Shevereagh, round which so many of them centre. “Since the Tuatha Dé seized the soil of Fotla, above the ranks of the noble druids in general is the branch of Cenn Febrat.’” ‘They were attached to the sacred mounds and tombs of Cush and Ballinvreena, where the chief gods and heroes of Corca Laegde “ Ernai” dwelt, or were buried, at Temair Erann. This treasure of legend was common property of the tribes of Dairfhine and Dergthene, and passed to their bards and_ historians. Whether some of it came from the chief religious centre of the latter tribe, the home of the great god Bodb, on Shevenaman, we cannot distinguish ; but its proto-history begins with MacNiad of the Ernai, Conaire, and Mog Neid, and his son Hogan? “ Mog Nuadat,”’ in the second century, and ends abruptly with Hanna Airethech, just before the introduction of Christianity’—a very eloquent fact. We, too, are fortunate in having much of it in a form preceding the synchronizing euhemerists of the tenth century and their incorrigible successors. Now what do we gain in broadest outline from these tales and poems ? There is no chronology ; after the “servile revolt” so called (which is very possibly a shadow of some then recent Milesian invasion), we see the so-called “serf tribes ” gradually spreading their power in Connacht (the seed-bed of the Tara kings), the Boyne valley, and at Tara, farther south in Leinster and in south Tipperary. Ptolemy shows at least one British tribe, the Brigantes, settled at the south- west corner in Wexford.'’ The Dergthene tribes of Magh (Feimhin) Femen covering all western and southern Ireland, and new tribes only in southern Co. Tipperary stand alone wedged among the “ pre-Milesians ” in Munster; the Ulidian tribes and the Picts fill up the north-east angle of the island, from the Boyne northward. In south Munster the arrangements on Ptolemy’s map agree well enough with the earliest sagas. In the latter the Erann, or Ernai,’ occupy the south- '“ Metricai Dind Senchas” (ed. Gwynn, Todd Lecture Series), x, p. 230. * In Gaul ‘‘ Hugen” is equated with ‘“‘ Esugen,” son of Esus. Cf. ‘‘ Les Celtes,’’ p. 30. 3 There were Christian settlements on the south coast, at Cape Clear and Ardmore, and at Saighir Cairain, before St. Patrick. The alleged monotheism of Cormac mac Airt is plausible, as he introduced water-mills, a trained army, and a large oblong hall, and so sympathized with Roman culture. * Were these the Siol Breogain, whose name suggested to Ptolemy the British tribe name ? 5 Ptolemy’s ‘‘ Iouernoi,” from whom Ireland seems to have been named. Cf. Ernai, Terne, Iouernoi, Iuverni. As I shall endeavour to show, ‘“ Ernai’ is a vague term like “Firbolg” or Milesian. In early literature it seems to include the Corea Laegde, Clann Dedad and the Muscraige, and Corea baiscinn. 132 Proceedings of the Royal Irish Academy. west from the “Genann” tribes at the Shannon mouth; the middle is held by the Muscraige; above them lie the Mairtene; above these again the Uaithne and Arada, north of the Saimer or Morningstar River. So far the arrangements in the Tain bo Cualnge persist, but a new tribe has come into sight in Femen. If its tribal genealogies may be trusted, it is a branch of that archaic tribe, living between the Brigantes and the Ernai, which called its chiefs “ Maqi mucoi Netasegamonas,” and recalled Nia Segamain, a great king’s name’ and little else, in the past, before Cuchulaind was born. This may suggest a British origin, for there seems no other trace of the war god Segomo in Ireland; perhaps they passed up the Blackwater valley from Ardmore’ to Seskinan, where their monuments are found, or by the great ancient road, the Rian bo, up the passes between the Knockmeildons and Comeraghs, into the Slievenaman footplains.* They were a feeble folk, under their chief, Mog Neid, and suffered from the might of the Ernai; but Hogan Mog Nuadat, probably by alliances or foreign weapons, asserted himself. He was believed to have divided Ireland with Conn, the King of Tara; at least “Conn’s Half” and “Mogh’s Half” were so explained; at most a nominal question of spheres of interest was touched between the “fair races.” In the tale, he and his father fell in battle, and his tribe was pitifully overthrown. His son Oilioll Aulom‘ was more fortunate (the alternative succession of the Dairfhine and Dergthene is probably as mythical as that of the Eoganachta and Dal Cais), but he appears as closely allied with the Dairfhine and fosterer of their prince Lugaid. Ile outrages the sanctuary of the Mairtene of Knockainey,* joins the land between it and the hills to Cenn Febrat, where he makes Dun Claire, above which he is buried, at the opposite side to the cemetery of the Ernai at Temair Erann. As his son Cormac was buried at Duntrileague, this may imply that the great cemetery of his race,’Oenach Culi® or Clochair, was founded later perhaps by Fiacha.’ Oilioll and the Ernai go ' “Of the Siabra’”’; his pretended chronology runs from about B.c. 150 to 50. He appears in all the Dergthene Pedigrees, though elsewhere very variant. * Ardmore is in Decies ; if (as seems established) the Dal Cais were ‘‘ the northern Deis” (B. of Leinster, f. 319¢ 5; B. of Ballymote, 171b 39; B. of Lecan, 174b 10), this is a remarkable coincidence that both people of the Deis were Magimucoi Neta Segamonas. The Eoghanacht prince in a.p. 630 is ‘‘ King of the Deisi of Magh Femen” (Eriu, v, p. 35). 5 ** Battle of Magh Leana ” (tr. E. Curry, Celtic Soc.), p. 1. * Supra, xxxiii, p. 479. ® Supra, xxxiv, p. 50. © A territory had a right to establish an 6enach (Ancient Laws of Ireland, v, p. 484). An 6enach was marked by a cladh (ibid., iv, p. 221). For Irish kings’ priestly functions see Rey. Celt., vi, p. 168. * Fiacha Muillethan appears in one legend as a northern king, if it be not another of the same name. Westropp—The Earthworks, &c., of S. L. Co. Limerick. 188 to war; after alternate successes he is left, old and feeble, uninjured, living to see the victor’s ruin and death, Now this story (whatever be its truth) is very remarkable. It evidently preceded the firm supremacy of the Dergthene apparent at the dawn of history. It is impossible to fancy court poets of either them or their rivals celebrating the overthrow and gross errors and follies of their princes, the dotage of Oilioll, or the ruin of Mog Nuadat and Lugaid mac Con, unless both tribes regarded the unpalatable story as beyond impeachment, “the memory of man going not to the contrary.” The subsequent slow advance of the Dergthene and quietude of the Ernai in the third century tells of exhaustion from war. The late “discovery” of their later sanctuary, Sid Druim (or Cashel), some say by King Core, son of Lugaid, in the fourth century, marks their slow advance. As its name implies, it was “the abode of idolatry and druidry ” (like Tara) when the idols fell on their faces at St. Patrick’s approach.’ The great warrior kings, Lugaid Meann and Conall, similarly, are followed by the long obseurity of the Dal Cais during the fifth and sixth centuries, when the Dergthene tribe had split into two, and the lesser tribe, the Dal Cais, was struggling to hold its own against Connacht until a.p. 620. Oilioll’s son, Cormac Cass, was mortally wounded in battle at Knocksauna, evidently trying to extend his territory westward. A century later all the east side of the Maigue valley was held, up to Carn Fhearadaig or Carnarry.*, When Lugaid Meann’s conquests were consolidated, we see Dal Cais colonies up central Co. Clare to Inchiquin Lake and the present north border of Co, Clare, a limit never exceeded. The tribes of the Tuath Echtge, the Corea Modruad, the Corea Baiscinn, the Aradha and Uaithne and the Ui Fidgeinte were allies, or allowed some nominal suzerainty to the tribes of Cass.’ Of course in the legends there is a great admixture of frankly mythical elements ; still these are in essentials common to all folk stories, and sometimes attach to late historic persons (even to Garret, Earl of Desmond, and Oliver Cromwell) in the same region. Ethnology has advanced too little 1 Ancient Laws of Ireland,” i, pp. 22-25; v, pp. 472-5. ‘Tripartite Life of St. Patrick,” pp. 41, 195. Cashel, founded by Core, see ms. R. I. Acad., 28D 5, p. 99. 2 Despite the usage of ‘‘ Carn Feradaig” for Carnarry till after 1550, and its always being defined as on the northern border of older Thomond, O'Donovan places it on the southern border on the authority of a chaotic poem naming it, Slieve Claire, ete., in no order, It was seized by Tigernmais from Conmael after the battle of Oenach Macha. Feredach, son of Rochorp, was buried in its cairn (Metr. Dind., x, p. 267), of which the base remains (North Munster Archaeol. Soc., i, p. 168, P. J. Lynch, and Proe. R. T. A., xxvi, p. 88). 3 The chronology is usually—Oilioll Aulom, A.p. 190-280 ; Cormac Cass and Fiacha Muillethan, mid. third century; Mogh Corb, 384; Lugaid Meann, 350; Conall, 378 ; probably the first authentic date in their history. 154 Proceedings of the Royal Irish Academy. to help us. All that is clear is that, unlike the dark “ Fir Bolg races,” the two Cashel tribes were fair, ruddy, large men, with golden, or red, hair, and blue, or green, eyes; the same is told of the Tara tribes.’ The legends, unlike the Red Branch ones, belong to the “ horse-riding period.” Much may yet be done by our scientific colleagues to solve racial questions ; till then, archaeology alone.can do but little. Myths are to be expected in early tales; the goddess Edaoin, making her favourite, Mog Nuadat, appear like a stone pillar while his enemies hacked actual rocks and pillars; Oilioll’s ear bitten off by the outraged goddess of the Mairtene; Cormac Cass and his singular medical treatment, can easily be classed, but a few other more intrusive matters in the more historic parts call for a note. I am not ready to believe that the mere name “ Lugaid” marks a god; “Lugodeccas” (of Lugaid) is common on the ogmic pillars, and Lugaid in the later history ; “ Nuada,” on the contrary, nearly always conno- tates the idea of “ the Severn god,” Nuada Argetlamh. There is a legendary “TLugaid Lamhdearg” in the chaotic past of the Dal Cais descent ;* the epithet is transferred to Lugaid Meann, on the edge of the historic period ; the pride of Munster and opprobrium of Connacht, he is entirely devoid of more super- natural features than attach to most favourite conquerors. Lugaid macCon, also, though in a mythic past, seemsa mere mortal ; as for his epithet we have “Cunogenus ” in Gaulish inscriptions, and “ MacCon ” among the descendants of his stepfather, the MacNamaras and others, to this day. With his sons, however, we have pure god myth, though, I think, rather late, when the triad was infecting everything ; their names are Oendia, Caindia, Trendia‘ (one god, gentle god, mighty god), but they are evidently intruded into some of the king lists (not all) merely to assert the poisonous myth of the alternate succession, for which history (even in the tenth century) was falsified and warped! “ Lugaid Delbaeth” seems an old god,‘ returned too late to earth. 1 Cormac mac Airt had “ golden hair, like blue bells were his eyes” ; ‘‘ his eyes were like (the bloom of) the sloe ” (Irische Texte, Ser. i, p. 204). The Dal Cais—‘‘ Lachtna, a fair man from Cragliath” (Book of Munster, circa a.p. 847) ; ‘‘ Dalchais of the yellow hair” (circa a.p. 900, Book of Rights, p. 81); ‘‘fair-skinned” (Wars of the Gaedhil, p. 79); and many allusions to their red and golden hair and green or blue eyes, in Cathreim Thoirdealbaith ‘from 1240-1318). * “ Magh Leana,” pp. 30-31. 3 The Cedraige, evidently from their name non-Milesians, are stated to be a branch of the Dal Cais (Battle of Magh Leana, p. 173). ‘New Ireland Review, xxvi, p. 135. So the three sons of the Boyne goddess are ‘*Tear-bringer, Smile-bringer, and Sleep-bringer ” (Rev. Celt., xxiv, p. 270) ; and the gate-keepers of Tara are ‘‘ Key, Post, and Valve” (ibid., xxii, p. 309). 5 As by the alleged “‘ Dal Cais kings” of Cashel, Aedh, 573, and Lorcan, son of Conligan. (See supra, xxxiii, p. 450.) ® **Delbaeth, son of Nét,” the war god (Leab. Gabhal, p. 153) Wesrropp—The Earthworks, &§¢., of S. BE. Co. Limerick. 135 He kindles five streams of magic fire, from which his sons, the fathers of the Delbna tribes, spring,’ and they are affiliated to the Dal Cais, undoubtedly without old warrant, for a mere political end. The horrible, and equally archaic, legend? of the worm of Cian seems transferred from the god Cian to the human eponymus of the Cianachta, who had also to be affiliated to the Dal Cais. These show that, under political stress, tribal pedigrees, like the later “Roll of Battle Abbey,” had a caddis-worm-like tendency to attach extraneous rubbish to themselves. Lastly, the simple churchman of the pre-Norse times (who tolerated god tales, and, at most, tried “to lure them into decency”) was replaced two centuries after by a priest full of the eviland wrongs of Norse heathenism, so gods had to become human—kings, wizards, jesters—anything but “gods.” The old god reappeared at different intervals in archaic literature, so the later reviser made him not one man but many. It is with these transformed gods and their local cultus that I am most concerned in the first half of this paper, as I am more concerned with the tribes and their legends and history in the second. Both studies are needed to elucidate the character of the earthworks (residential or sepulchral) of the ancient Clit mhail mhie Ugaine in south- eastern Co, Limerick.* (B)—TuHE Gops IN Criv. We ventured to embark on the dark and stormy sea of Irish mythology to search for the uncharted subject of the local sanctuaries in Clin at Cush, Knockainey, and Clogher. We established the objects of our search; but a mass of material of equal importance remains about other places and gods,‘ both on the ground and in our literary sources, so we cannot leave it aside, but must deal with, and leave it to the amendment of the students of its obseure and difficult subject. It takes courage to attempt to deal with the Irish gods. The very word causes a nervous feeling that one’s work might get confused with the work of the older antiquaries, who sat and created, without study, visions of Pheenician, Cuthite, Indian, serpent-, cow-, and even pig-gods worshipped 1 New Iv. Reyv., xxvi, p. 130 ; xxv, p. 73. ? Hib. Lect., iv, p. 392. 3 Clfu and Luchair evidently overlapped in the debatable land of Coshlea, Luchair extending to the east of Cenn Febrat and to Tul Tuinne, Dun Tultha (or Tonntinna), on Loch Derg (Metr. Dind S., x, p. 239). Glenlara, named in the ‘‘ Battle of Magh Leana” as apparently near Killarney, is more probably Glenlara on Cenn Febrat, to which a defeated prince flying from Magh Feimhin to the south coast at Beare might more naturally have sought temporary refuge among its tangled oak woods and streams. The territory of Curoi mac Daire extended between Knockainey and Slievereagh (Mesca Ulad, p. 17). 4 Supra, xxxiv, p. 53. R.I.A. PROC., VOL. XXXIV, SECT. C. [21] 156 Proceedings of the Royal Irish Academy. in Ireland, or with that of inventors of druidical rituals and arrangements of earthworks to form eagles or serpents’ on the map of Ireland! One must prepare for bitter hostility, not only from the unexpected survivals of these discredited theorists, but from students of the old schools, whose faith in the pseudo-chronology and pseudo-history scientific antiquaries must endeavour to demolish. I will avoid the terms “fairy” and “elf” for the Sid folk, for each suggests the English idea, and gives as false an idea of the hero-like Sid as could well be imagined. The terms “demon” and “goblin” are, perhaps, even more misleading. There is no necessary implication of ugliness, or wicked- ness and malignity, in “the bright gods of elder time.” The Badb was usually beautiful and stately—a Pallas or Artemis. She only gradually reaches the nadir of loathsomeness and horror—“ nor uglier follow the night- hag” than she—in late works, culminating in the Cathreitm Thoirdealbaigh. Even to the present day some goddesses—Aibinn, Aine, aud Cliodna, for example—kept all their supernatural beauty. There were two main groups of gods in old Ireland,* the Tuatha Dé Danann (many of whom are traceable in Gaul and Britain) and the Fomorigh. As in other primitive faiths, the “departmental” allocation of divine functions (elaborated by the later keen Greek intellect, and accepted in globo by the unimaginative, practical Roman) was not followed by the Celt. The gods overlapped each other in attributes and functions. Lucian was irritated and scandalized at the mixture of Hercules and Mercury in Ogmius in Gaul; the god Nodens seems to have resembled Mars and Neptune, with, perhaps, a trace of Apollo and Pluto, so we cannot take absolutely the identifications (even when made by their worshippers) with the Roman gods. I find no trace of the other family of gods ontside Ireland, nor, indeed, in Cliu, unless, perhaps, of Bress; and even he is a half-breed. Besides these, however (and the Aine family was of the Tuatha Dé), area number of gods of other races, often, possibly, non-Celtic, from the great Bodb Derg to the obscure mountain-gods, like Cliu, Donn, and Febra; and ancestral ones, like Cairbre Muse, Dergthene, and Deda, with his forty sons. The subject in its crude beginning leaves no excuse for dogmatism. 1 I know the possessor of this remarkable map, but not its author. * For the general subject consult various articles on the Gaulish and Irish gods in “Revue Celtique”; Hib. Lect., iv; ‘‘Ir. Myth. Cycle”; Arthur B. Cook, “The European Sky-God” (Folk-Lore, xvii, p. 28); “ Encyc. Relig.,” iii; Charles Squire, ‘“Mythology of the British Isles”; Dr. J. A. MacCulloch, ‘‘ Religion of the Ancient Celts.” Wesrropp—Vhe Larthworks, §¢., of S. 2. Co. Limerick. 137 I state what seems to me most probable views; but some, at least, must get overturned by further research. Though we have a substantial mass of early statements, even good, but slightly later, authorities do all they can to contra- dict the obvious facts, and make the Tuatha Dé not “demons or fairies, but descendants of Japhet.” All such euhemerist assertions are worth nothing. The “Fomorigh,” or “under-sea folk,” were probably gods of non-Milesian tribes. Only in the ranks of the Tuatha Dé do the gods of Gaul and Britain appear. The “Fomoraig” are “champions of the Sid,” probably of the holy mounds, not of their rival gods. We have a similar phrase in “the host round the cairns,” in St. Columba’s poem, and the Coir Anmann says that Cormac Coinloinges, “who watched at cairns,’ was called Mia in Chain, “the champion of the cairn.” Of the Tuatha Dé we have full evidence from Gaul of the worship of Lug, Ogma, Neman, Net, Danu, Grian, Brigid, and Anu, with, perhaps, Midi” and Bile. Along with them we have the three crow-like goddesses of war, one called Bodbh of Battle, “(C)athubodua,”’ and the divine bull of the legend of Cualnge, “ Donnotaurus.”* The worship of many of these gods in Britain is equally well established. Lleu or Lugh; Nudd (or Ludd) Lamereint, Nuada Argetlamh; Manawyddan, son of Llyr (or Manannan); Don or Danu ; and Govannan or Goibniu. The sacred mounds were divided between the divine races by Manannan and Bodb Derg; but they were “people of the Sid” (whatever exactly was intended), probably long before our literary sources begin. Traces of the worship of the following gods I have so far found in Clin :— Lug, Nuada, Nechtan, Cuil, Aine, Ferfi, Segomo, Net, and Manannan, of the Tuatha Dé. Of the non-Milesian gods I may note Bodb Dearg, Cliu, Daivine, Dergthene, Deda, Garban, and Febra, Donn Firinne, Edaoin, Cliodna, and a swarm of the relations of Deda, son of Sen, including the mountain goddesses Mis, Echtge, and Eblinn, and the river goddess Sinann, ' Loc. cit., p. 403. ? This I doubt, as the bull and Phrygian cap suggest that ‘‘ Medros” is Mithra (Rev. Celt., xxv, p. 47). The question whether Berekynthia and Brigendo are the Ivish war goddess Brigid deserves more study. I venture to suggest that the mortal representatives of certain gods bore their patrons’ names; the priestess of Knockainey may have been “‘Sine.” I do not think (as has been suggested) that kings were called ‘‘ Nuada”’ ; the name at different periods arises from chronologers dating the same god at various generations. Lastly, the equation of Bile with Balor rests oniy on similar but doubtful pedigrees. ’ Borne by human chiefs, as by the prince of Helvii. For Tarvus Trigaranos see Rey. Celt., xvili, pp. 255-4. [21] 138 Proceedings of the Royal Irish Academy. (C)—Lirerary TRACES OF THE GoDs. It is a widespread commonplace in Ireland that our ancient literature Says nothing about the early gods. This arises from the acceptance of the euhemerist statements, and, like them, is unfounded.'’ The pagan gods are fully recognized as such till the late tenth or even mid eleventh century. High clerics, like Cormac, felt as little hesitation to name them as St. Luke felt when he named Zeus, Hermes, Ares, the Dioscuroi, and Artemis. Evidently the new prejudice roused in the wars of the Norse and Danes rendered it desirable to sweep away every trace of a pagan deity, though the survival of so much ancient literature rendered the effort unsuccessful. Let us trace this decline from the frank statements of the early writers. Tain bo Cualnge (edited and written down in the early seventh century, and, in its existing form, as old as the eighth-century glosses) was fragmentary when first transcribed. It tells of the direct intervention of Lug, Neman, Manannan mac Lir, the Bodbh, and the Morrigu, daughter of Ernmas, and it regards Cu Chulaind as the son and re-incarnation of the sun-god Lug himself.* Phrases like “the blessing of gods and non-gods” occur in the archaic Fled Bricrend.‘ Even the “godless” Tain be Flidais has an appari- tion of the war spirit. JLeabar na hUidre plainly calls Conchobar and Dechtire gods and earth-gods ; and Dechtire’s son, Cu Chulaind, is a god, descended from the gods Net, Elathan, the Dagda, Cermait, Lir, and MacCuill, the husband of Ireland herself. A passage from the very early “Yellow Book of Slane” speaks of Cu Chulaind’s holy mound, Sid Setanta, in Muirthemne.* - Tirechan’ (A.D. 656) tells how St. Patrick and his com- 1 This feeling occurs in ** Book of Ballymote,” Sdn, 31,344, 39. See the Leabhar Gab- hala, pp. 141-189, for list of Tuatha Dé, though it has clerical colouring, such as their wars along with the Athenians against the Philistines ; also ibid., p. 163 ; ‘‘ Battle of ae Tured”’ (Eriu, viii, pp. 17, 35, 45); and the late ‘‘ Cath Finntraga ” (ed. Meyer), . 40. 5 * The salient works are—‘‘ Tain bo Cualnge” (c. 630) ; ‘‘Sanas Chormaic” (c. 896) ; Eochaid Ua Flainn. For the later phases—Cinaed Ua Articain (974); Cuan Ua — (1024) ; Flann of Monaster (1050) ; Giolla Coemhain (1070). “Tain Bo Cualnge” (ed. J. Dunn), pp. 90, 91, 160, 347 ; the Morrigu, p. 175 ; Lugh, 182-3 ; Badbh, 27, 29, 188, 191. Manannan, 190: some euhemerist interpolates ‘‘ The Tuatha Dé considered their men of learning to be gods and their husbandmen non-gods.” Cf. Leab. Gabh., 143. ‘* Badbh’s corpse-fold,”’ p. 188, is a pile of bodies; ‘* Torches of Badbh ’’ are weapons. *“ Fled Bricrend ” (Ir. Texts Soc., ii, pp. 57, 61, 67, 95) in a ninth-century recension. “ Death of Cu Chulaind ”’ (Rey. Celt, iii, p. 175, and us. T.C.D., H. 2-27). We hear of Cu Chulaind’s ** Folks of might whom he worshipped.” * Leabar na hUidre, p. 101b, Book of Leinster, p. 123b. ***Sick Bed of Cu Chulaind” (Atlantis, i. p. 390,; also ‘‘ Heroic Romances of Ireland,” A. H. Leahy, No. 2, p. 65; Da Derga’s Hostel (Rev. Celt., xxii, p. 66). * Published in ‘‘ Tripart. Life,” vol. ii, p. 315. Wustropp—The Lurthworks, §e., of S. bk. Co, Limerick. 189 panions were mistaken for “men of the Side, or earth gods,” and were asked, “‘ Are you of the Sidh or of the gods?” Note how little a Christian writer then hesitated to tell facts; but Tirechan’s candour is as nothing to that of Cormac, King-Bishop of Cashel (about A.D, 896). In the latter’s Glossary! we find—* Art, a god”; “Ana, mother of the Irish gods”; “Brigid, a goddess worshipped by poets”; “Diancecht, god of the powers, god of healing” ; ‘Manannan, the Irish and British call him god of the sea”; “ Neit, a god of battle with the pagans of Gaul”; “Nemon, his wife”; “the three gods of poetry”; and he often refers to sanctuaries and magic rites. He calls his predecessor, Hogan, “ Mog, son of Nuada,” and mentions ‘Aine, daughter of Kogabal. Coir Anmann* has many early notes—“In Munster was wor- shipped the goddess of prosperity, whose name is Ana, and from her was named the ‘Two Breasts of Ana,’ over Luachair.”® It mentions the gods Nuada Derg Lamh, “Nia Segamain of the Siabra’ of Engus; Nuada Necht; Nuada Finn; Nuada Argetlamh; beside more definite entries—“ Manannan mae Lir, god of the sea”; “the Dagda, the ) ; Nuada Salfada, son fire of god; he was a beautiful god, for the Tuatha Dé Danann worshipped him, for he was an earth god to them,’ and “ Diancecht, deus Salutis.” Even the rather late “Battle of Magh Leana” (p. 90) tells of Conn’s “powerful friends,” the Tuatha Dé, helping him in that combat, and the Mesca Ulad names “the great Dagda, son of Ethliu, the good god of the Tuatha Dé,” with “Delbath, son of Ethliu,” Aengus Og, son of the Dagda and Cermait honeymouth, “three noble youths of the Tuatha Dé.” In Latin bo Cualnge, Bruden Da Derga, and Hed Bricrend abound many allusions to pagan gods'—“TI swear by the gods by whom my people swear,” “by the god of my people,” or “of my tribes” or “by whom the Ulster folk swear,” “by the gods whom I adore,” and “we give thanks to the gods for our return to Erin.” Indeed it was impossible to expurgate the old paganism and its phraseology. They even got copied into late works like the Agallamh, which, despite its sturdy euhemerism, says—“ the Tuatha Dé Danann, who are unfading and whose duration is perennial.”? The common sense of the old scribe of the Zain, in the Book of Leinster, looked facts in the face, but did not omit them—*I, who have transcribed this history, or rather fable, do not 1 «¢Sanas Chormaic”’ (ed. W. Stokes), pp. 3, 4, 17, 23, 47, 56, 67, 90, 99, 114, 122, 145. 2 Trische Texte, Ser. iii, 12, 1897, pp. 289, 291, 295, 327, 355, and 3d7. 3 Also note Doonass (Hasa Danainne) rapids on the Shannon, 4A curious outcome remains of euhemerist attempts to clear away gods like Lugh, MacCecht, &e., by giving them other gods. 5 Mesea Ulad, pp. 31-33. 6 Even if euphemisms for a divine personal name, they still ‘‘ call gods ‘gods.’ ’ 7 Agallamh ii, p. 208. 140 Proceedings of the Royal Irish Academy. believe some things (in it), for some are delusions of demons, some are poetic figments, some seem true and are not.” It is curious to find some belief in the Sid folk in even so orthodox a person as Rey. Geoffrey Keating, about 1630, where he does not understand how the euhemerists obtained tidings of such early events and wonders: was it aerial demons, the Sid lovers of the oldest inhabitants, who told it? or was it engraved on flags of stone ? The gods so recorded are of the Aryan type, not the grotesque fiends of America and Asia, the gloomy. but impressive, gods of the Euphrates and the Nile, but supermen, like those of Olympus or of the Valhalla; “departmental gods,” not almighty, subject to fate, wounds, mutilation, and pain, and the joys, passions, and sorrows that beset mankind. Sometimes great nature powers, like the Sun and the Ocean, they were yet compelled at times to seek human aid, and were parents of men by mortal lovers and protected,*though they could not always save, their mortal offspring or favourites, as Zeus could not save Sarpedon. Their tale continued to be written and (when at last it passed into the hands of credulous pedants, bent on giving Ireland a prehistoric history, at least to the days of Noah) it was an invaluable quarry from whose fragments mosaics of false history and imaginary king lists could be made. These waifs and wreckage (washed up from a past ten or twelve centuries before the writers) became, not the exquisite stories of Greece or the fine sturdy tales, full of humour and pathos, of the Edda, but dried and mutilated husks, duplicated, split up,’ or blended with others, set in contradictory frameworks of chronology, where a god, like Nuada, becomes various persons in the year before our era 1741, 1007, 600, 150, 112, or 50, and the Milesian invasion took place in B.c. 1569, 1229, 1066-71, 554, or 331. There are yet persons living who regard any criticism of this turgid mass of perverted compilation as blasphemy against the greatest glory of our land, but their time is past, and soon the pre-Christian annals will be ranked with the tales of Geoffry of Monmouth, and left to mere country archaeologists of the older schools. Then, when we have removed the strata of euhemerist fiction and rubbish? from the ruin, the foundations and beautiful fragments of the once noble fane of Irish mythology will stand clear to the sun. Doubtless when that is done men will wonder, no less at the large amount which has escaped the destruction of over fifteen hostile centuries than at the credulity that so long held, against all reason and evidence, the mass of late fiction which concealed it to be the only true history of early Ireland. | Like the divergent versions of the Battle of Magh Tured. 2 «‘ Rubbish,” like the pre-deluge legend, Keating, ‘‘ History,” i, sect. v. Wesrropp—The Eurthworks, §e., of S. EB. Co. Limerick. 141 . (D)—LocaL Cunrus or THE GREAT GopDs. Lue Lamnrapa. Most glorious of all the gods of Erin is “Lug of the long hand,” “the god, twin-born with the day,” the sun. His epithet was a note in many a religion, the creed of every high religion that “God’s hand was not shortened,” and recalls the pictures of the Dise God of Egypt with his hand-ended rays blessing his devotees. The euhemerists could not conceal Lug’s nature—< Lug, like the sun is the splendour of his face, men are unable to look upon it”; “as brilliant as a summer’s day he rose from Manannan’s territory in the east”; “he rode the steed of Manannan (the white-maned wave), swift as the bleak, cold wind in spring” ; when he is in the west men ask “what else than the Sun is it? It is the radiance of Lug Lamhfada.’” He was “ Master of all the arts,’ “ Lug, with whom are all the arts.” Caesar when he spoke of the Gaulish Mercury, “the inventor of all the arts,”* evi- dently meant Zugus. In Gaul he was a centre of cultus; the towns called “Tugdunum ”* were his special seats, three still echo his name—Lyons, Laon, and Leyden; the fourth is now St. Bertrand de Comminges, where the Lugnasad festival was kept in August, as it was in Ireland. He seems to have been personified (like the war goddesses) by a raven, “ Lougos,” being so translated ;° so Odin had two raven spies. He and his divine “ boy,” like Lug and Cu Chulaind,® were represented in Gaulish carvings. Not to repeat his long story in Ireland, he was son of the Dagda, but his mother was Ethniu, daughter of the horrible darkness god Balor, the god of the evil eye. Lug’s slingstone drove the eye through the fiend’s head, as his Greek equivalent Hermes slew Argos, the many-eyed night. He was worshipped by Mac Greine, one of the three divine husbands of Erin. Men were called from him Mog Loga; his great sanctuaries were at “ Lugmagh,” or Louth, and Naas, which last was known as “ Lis Logha”; perhaps its mote 1 Hib. Lect. iv., p. 384; ‘Ir. Myth. Cycle,”’ ch. xiii. 2FWate of Children of Tuireann” (Atlantis, iv, p. 161); ‘‘Fets tige Chonain,”’ Ossianic Soc.), p. 2d. 3 Rennes Dind. S. (Rev. Celt., xvi, p. 77), ‘‘ Irish Nennius,” p. 47. 4 Hib. Lect. iv, p. 419. A fifth Lugudunum belonged to the church of Le Mans. “‘Tugnasad,” see Sanas Chormaic, p. 99, Hib. Lect. iv, pp. 418, 419; but, to contrary, Rennes Dind. 8., Rev. Celt., xvi, p. 51. 5 Folk-Lore, xvii, p. 164, ravens appeared in flocks when Lugdunum was built. Manuel de l’Antiquité Celtique (Dottin, 1906), p. G4. 6L. na hU,, f. L01b ; ‘The Cuchullin Saga ” (Miss Eleanor Hull), lvi, lxii, pp. 15-20 ; ““Compert Con Chulaind” (Rey. Celt., ix, pp. 1-15; ** Tain bo Cualnge,” p 96. Cuchullain and Lug were too individualized to merge into one. For Gaulish carvings see Rey. Celt., xxvii, p. 319; xxviii, p. 224, ‘“‘ Lug and Cu Chulainn, his son, in Gaul.” 7 “Tr, Myth. Cycle,” p. 113. 142 Proceedings of the Royai Irish Academy. was at first his Si? mound.! He also dug the great earthworks of Tailti, at Telltown, Co. Meath. The rainbow was his “hurling-stick,” and the Milky- way his “chain.” In late times the annalists made him a king, from A.M. 3370 | to 3331 (B.c. 1871-1830, but others said 1714-1674), and writers put in his mouth the degrading confession, “I am Lug Mac Ceithlenn, of Adam’s race, who have returned from the dead,”* making him ride a charger as a late feature. Some said his valour was only a seventh of that of Hector, but was (even so) 2401 times that of the bravest “modern hero,’ Murchad, son of Brian, in 1014! He was, however, cited as an example to fight against the Danes, for he “sprang over every obstacle and exterminated and expelled the foreigners and pirates out of Ireland.”* Even so late as 1750, Seaghan O Tuama compares Prince Charles Edward to Lug of the trenchant sword.’ Lug was ancestor of the Corea Laegde,* and grandfather of Conaire (whose son, Cairbre Muse, we shall note); his descendants lay in the cemetery of the Ernai or Slievereagh, and perhaps one of the three tumuli on which “men feared to sit because of the Tuatha Dé” was dedicated to the sun god. He was also divine father of the Corca Oiche and Corcamodruad.? The festival of the goddess Aine at Knockainey depended on Lug’s festival, and was cele- brated for three days after, and continuously with, it,° for she being a harvest and meadow goddess, and giver of fertility to crops and cattle, naturally needed the sun. It will be remembered that Art Imlech gave his name to Emlygrennan, below Slievereagh, and dug the forts there; he was slain by Bress. The latter had been taken captive when Lug slew Balor and warriors of the Fomoraig “as numerous as the son of Ler’s horses in a sea storm.” To save his life, Bress offered that the cattle should give milk all the year round, and that there should be a harvest every quarter, but, only when he told his captors the lucky day for harvesting, was he set free.’ In another story, when Nechtan, son of the god of Nuada, reigned in Munster (his wife’s sanctuary was at Clogher, near Emlygrennan), Bress imposed on him as tribute the milk of 100 dun cows for every house. “Lugh, who was dutiful on all occa- sions,” helped the overtaxed province. He singed and dyed all the cattle to ' Folk-Lore, xvii, p. 157 ; ‘‘ Manuscript Materials,” p. 478. ? Squire, p. 62. ***Tr, Myth. Cycle,’’ p. 171; ‘* Manuscript Mat.,’’ p. 388, later than 1014. ‘“ Wars Gaedhil,’’ pp. 187, 188. 5 « New Ir. Reyv.,” v, p. 38. ® So in ** Corea Laidhe,’’ p. 9, p. 57, which makes him ancestor of the Dergthene also. B. of M. Leana, p. 172. 7“ New Ir. Rey.,” xxvi; ‘‘ Ancient Ir. Genealogies” (J. MacNeill), pp. 132-4. ° R.I. Antt., Ir., ii, p. 36. ® “ Voyage of Bran,”’ ii, p. 178. Wesrropp—Vhe Larthworks, §¢., of S. £. Co. Limerick, 143 dun colour, and made sham cows full of unwholesome fluid, and, as Bress was under a gets to drink the milk, the oppressor got poisoned and was buried under Carn Ui Neit (called from him as descendant of the war god Net) on Mizen Head, Co. Cork.! NUADA AND THE War Gops.—The connexion between the god Nuada and the ruling races and districts of mid-Munster is well established and of much mythological importance. Like most primitive people, the Celt eloried in the divine descent of his tribe and chiefs. Professor MacNeill is most pro- bably right in regarding the ancient accepted genealogies as “ ending” with (1e., leading up to) a god.* Early people claimed not only the personified nature power, but the super-man and super-beast as ancestors. The Gaulish inscriptions give us names like Bodvogenus, Camulogenus, Esugenus, Totatigenus, Morigenus (“ Morgan ”’), and Rhenogenus,’ and the general term Devogenus (God’s son), as the man derived himself from Bodb, Camulos, Esus, Totates, the Sea or the Rhine; while others bore animal names, like Cunogenus (MacCon), Urogenus, and Artiogenus (MacMathgamhain), the non-Milesian Trish tribes Conraige (hound), Sograige (bitch), Gamanraige (calf), and Osraige (fawn, or even ofso,, werewolf)! attest the other phase.’ I long hesitated to receive the names in genealogies as actual gods (regard- ing them as mortal adaptations, like Diogenes, Phoebe, and Silvanus); but the presence of Lug Lamhfada, Manannan mac Jir, or Oirbsen, and Nuada Argetlamh is unmistakable. Nuada was closely bound up with Munster— Jathaoir Mor (ancestor olf the Dairfhine, whose tribal ancestor Lugaid Laegde, was buried at 'emair Erann on Slievereagh), descended from Nuada Necht and Nuada Finnfhail. The rival Dergthene not only called their first great prince “ Mog Nuadat,” servant of Nuada, but were derived from Nuada Argetlamh in direct descent; indeed, Mog Nuadat (af Cormac’s Glossary be correct) was a son, being called “ Mog, son of Nuada.’”® Now, in view of the prejudice against treating the late eleventh-century imaginary king lists as anything but history, I must first methodize my material to show the identity of Nuada under his various surnames. Nothing 1 Metr. Dind S., x, p. 217; see supra, vol. xxxii, p. 271. Hackett and Borlase regard Leaba Mo Laga as a shrine of Lug (Dolmeus of Ireland, iii, p. 769). * New Ir. Rev., xxvi, p. 134. The Milesian groups of descents converge respectively on Conn of Tara, Cathair Mor of Naas, and Oilioll Aulom (p. 7), and meet in the god Nuada. 3 Cf. ‘ Neilogenes.’’ New Ir. Rev., xxv, p. 7. For ‘‘ Trish werewolves’? see Encye. Relig., viii, p. 207 ; Rev. Celt., ii, p. 202. Wolf names like Faelchu, Faclad, &e., abound among the Ossorian chiefs. Nuada was ‘‘ Lord of the wolf,” and Bress was a ‘* wolf-man.”’ ®«« Anthropomorphism among the Celts,’’ De Jubainville (Rev. Celt., vol. xix, p. 229). 6 Of course, like Cu Chulaind and Mongan, he had a human father as well, R.1,A. PROC,, VOL, XXXIV, SECT, GC. (22] 144 Proceedings of the Royal Irish Academy. was more common than to divide a god so as to adapt him to the Triad formula. Brigid is triple, “of the smiths,” “of the poets,” and “of the doctors”; “Eriu, Fotla, and Banba are Ireland; MacCuil, MacCecht, and MacGreine, her husband ; Macha, Neman, and Bodb, the war goddesses, are interchangeable ;? while triads of separate gods were worshipped every- where, as “ Alldai, Buada, and Buan”; or “ Oengus, Boadan, and Elemair,” at the great Boyne Tumuli;* “Sinann, Dairine, and Macha,” at the Shannon ; “ Kogabal, ‘Aine, and Fer Fi,” at Knockainey ;* and “Nuada, Nechtan, and Boadan,” on Carbery Hill, above Trinity Well, the source of the Boyne. This well may have been so dedicated to conciliate those who worshipped the old divine triad, in order that, like Patrick, they might “ bind themselves to the Trinity.” The fertility of the Irish mind in inventing genealogies is exuberant. Even Bunyan, with his convincing parentage and biographies of personified vices, pales before the riches of [rish allegory. We meet endless characters like ‘‘ Violent, son of Extinction, son-of Darkness, son of Ailment ” (in the Dind Senchas of Carman); “Serutiny, son of Cogitation, son of Great Knowledge, son of Enquiry ” (in Cormac’s Glossary) ; “ Yellow, son of Fair,” and “Terror, son of Great Fear” (in Med Bricrenn); the Dagda’s wives, “Lie, Guile, and Disgrace,” and the druid’s three sons, “Prophecy, Knowledge, and Truth” (in the Battle of Magh Leana).* This teaches us caution in receiving genealogies as evidence for the non-identity of the Nuada names. Chronology proves it as little. The Nuadas are dated B.c. 1741, 1007, 160, 112, or 50; but then (as we saw) the Milesian invasion has the dates B.c. 1569, 1229, 1066, 1071, 554, and 331; Queen Medb’s father, Eochu, is dated 8.c. 137, 56,and 27 ; Eochu Mumho, eponymous King of Munster, who fell in Cliu at B.c. 1429, 1275, and 776. Endless other “ synchronisms”’ show that, so far from being “authorities,” the chronologers could not agree to tell the same story.’ Such systems are less authentic than the very myths ef the gods! and it must be remembered that the ever- living gods naturally appeared in different ages in stories which, when ‘Tr. Myth. Cycle, pp. 210, 218 ; Hib. Lect., iv, p. 579; Rev. Celt., vii, p. 283. 2 W. M. Hennessy, Rev. Celt., i, p. 33; Proc. R.I. Acad., x, p. 425. > Metr. Dimd S., x, p. 27. ' Silva Gad., vol. ii, p. 575. ® “ Fled Bricrenn,’* p. 9; ‘*Sanas Chormaic,”’ pp. 144-5 ; ‘‘ Magh Leana,”’ p. 155. ® A good example of duplication is in the O'Driscoll pedigree ; (57) Lugaid, son of (56) Daire, grandson of (54) Siothbolg, son of (53) Daire, son of (52) Siothbolg, (Keating, ivy, p. 115), and in the Eoghanacht pedigree three recurrences of the god Nuadha are noticeable, evidently independent documents (Book of Lecan, f. 64b). In fact, these pedigrees are heterogeneous fragments stuck together. ? The parentage of several of the great gods (e.g. Mider) is most variant. Ir. Myth. Cycle, p. 178n. Wesrropp—The Lurthworks, §¢., of S. 2. Co. Limerick. 145 “synchronized” and “ rationalized,’ compelled the same god to be “distributed” as different personages ages apart.! Indeed, divergence of name and epithet is no evidence for the non-identity of a god, or Zeus Dodonaeus, Olympius, Lykaeus, or Dictaeus ; Athene, Pallas, or Tritogeneia ; Thor, Yee, Bolvérk, Kialar, Sidgrani, or Hlorridi, should be so disintegrated. On this account, I think the theory that the various “ Nuadas” are reincarnations of the god is refuted by the evidence for identity, as I shall tabulate. TABLE ON IppeNiTITY oF NUADA. (1) Nodens, Nudens, Ludens, Nudd, or Llud.* An early British god of war, light, and rivers. He was worshipped on the Severn at Lydney, and on the Thames at Ludgate. In Ivish literature Loden, son of the sea god, was a great cattle-owner; so is Nudd in Welsh literature. The ancient name was Nudens lam argentios, according to Rhys, “ Ludens” being alliterative to Lam Ereint ov “ silver hand.” (2) Nuada Argetlamh, “silver hand,”’! king, and war god of the Tuatha Dé, a light god, son of Echtach, or Eochu. He is ancestor of the Eoganachta and Dal Cais princes.” Dwelt at Almhu, father of the god Tadg,’ succeeded by Bress. (8.c. 1741-1721.) (3) Nuada Fail, or Finnfail (‘‘ white light,” says Rhys), son of Giollehaid’ (cf. “Mochaid”’), Ancestor of the Dergthene. Slew Art Imlech of Cliu,® and was slain by Bress, who succeeded him.° (4) Nuada Derg, son of Achi(Echtach), or of Sedna Sithbaice,"” by Dairfhine, the ancestress of the Corea Laegde, Fosterer of Hogan, who was 1 Tt is necessary to repeat these facts for emphasis. * Hib. Lect., iv, p. 119; Ir. Myth. Cyc., pp. 86, 88, 118; Holder, ii, 754, Encyc. Relig., 1, p. 284. ‘Ludd Lamereint” appears in the story of Kilhweh and Ollwen. Nuada, Abbot of Armagh, a.p. 810, was also called Noda (Trip. Life, vol. i, p. 82). 3 Hib. Lect., p. 125 ; Eneye. Relig., p. 284; Hib. Lect., pp. 128, 290. ** The cattle of Nudd the bounteous were 20,002,” Folk-Lore, xvii. + Poem of St. Columba (Ossianic Soc., vy, p. 255); the epithet recalls a Pict, *¢ Argentocoxus,” or ‘‘ Silver Thigh’’ (Dion Cassius, Book I, xxvi). ® Book of Lecan, f. 64b; ‘‘Magh Leana,’’ pp. 170, 171; Keating, ‘‘ History,” iv, p. 17. ®Sanas Chormaic, p. 55; Leab. Gabh., p. 149; Atlantis, v, p. 259; Rey. Celt., i, p. 52. 7“ Magh Leana,’’ 170; see also ms. R, I. Acad., 28, K 32, p. 108. 8 Book of Leinster, Todd Lect. Ser., iii, p. 171. 9 «« Bress a wolfman’’ (Jour. R. Soc. Antt. In, xlvy, p. 38; Giolla Coemhain and Leab Gabhala, pp. 165, 159 ; supra, xxxiii, p. 47 ; ‘‘ Art’’ means a god (Sanas Chormaic, _ W Silva Gad., ii, p. 5, 9; also 3, 8, 189. Setna was king of Claire (Cenn Febrat) B.c. 909 ; see also “‘ Magh Leana,”’ p. xxi. ‘Second Battle of Magh Tured”’ (Rey. Celt., xii, p. 246). we {=4" | 146 Proceedings of the Royal Irish Academy. - thence called “ Mogh Nuadat.” Dwelt at Almhu;' a great landowner and fort-builder ; he hada son Glas (cf. N. Derg ILamh, No. 7), and was reverenced, with Nechtan and his wife Boand, at the source of the Boyne. Nuada, son of Achi, was husband of Almhu and father of Tadg. (5) Nuada Neacht, son of Setna Sithbaicc* (ef. Sithbale, grandson of Nuada, son of Lug, infra, No. 10, and Sithbolg, fifth from Dergthene, and sixth from Nuada Airgthech, in Keating). Husband of Boand, who herself met the fate of Sinann, daughter of Loden.* Ancestor of the Dairfhine; father of Tadg of Almhu,‘ slain in Cliu,* B.c, 118, builder, with “ Mog Nuadat,” of the fort of Almhu. (6) Nuada Fullon (the beautiful), a magician, caster of magic wisps, B.c. 600,° son of Allot (brother of Bress).7 Along with Nuada Finnfiail and Nuada Necht in tribal pedigree of the Dairfhine in the Book of Lecan, f. 64b. (7) Nuada Derg Lamh, father of Glas, an alias of Nuada Dearg, but dated a.p. 430. (8) Nuada Salfota, an alias of Nuada Dearg,® with same story of Mog Nuadat and the fort-making, but at Ailinn instead of Almhu. (9) Nuada Find (note epithets “ white,” “silver,” &e.), reared at Find Mag, or Mag Femen. (10) Nuada Aicnech (? Airgthech), Luigthine, son of I.ug, grandfather of Sithbale of the Dairfhine, and father of “ Daig Dergthene,” circa B.c. 50. (11) Nuadha Dearglamh, son of Kochu, third from Eber Finn, son of Bile, fourth from Alldod (Allot), son of Nuada.’° (12) Neachtan, son of Nuada Necht, husband of Boand, son of Labraid, or of Namat,'' a great cattle owner; “Cu Nuadat,” in one Boyne legend. He ' Rey. Celt., ii, p.84, from “‘ Battle of Cnucha.’’ For Almlu, Agallamh, ii, p. 132. No trace of a fort or cairn remains, but several of these elsewhere have been entirely obliterated (even in my memory), especially those on rocky knolls. * Sedna was grandfather of the hero Finn (Silva Gad., ii, p. 166, p. 99, and p. 519) ; see also Genealogy, Keating, iv, p. 42. Sedna is sometimes given as son of Bress (ms. R. I Acad., 23K 32, p 108). °Sinenn, Metr. Dind S., x, p. 27; ‘*Duanaire Finn,” p. 133; ‘‘ Boyne and Black- water "' (Sir W. Wilde), p. 34; Tain bo Cualnge,” p. 148 ; for Dun Mic Nechtan, see Book of Leinster, f. 65b. * Silva Gad.. ii, p. 519; ‘‘ Duanaire Finn,” p. 137. * Verse of Senchan Torpeist (c. 630, Todd Lect. Ser., xvi, “‘Fianagecht ”). ® Coir Anmann, Irische Texte (Stokes and Windisch), iti, p. 367. ? Leab. Gabh, p. 153, sons of Elathan, son of Dealbaoth. 5 Ibid., p. 305 ; “* Magh Leana,” p. xxi (from ms. T.C.D., Hs, 18 £567), and pp. 1, 117 5 both are called “‘ son of Dairine.”’ ® ** Fianaigecht '’ (Meyer, Todd Lect. Ser., xvi, p. 29). 1” Keating, Hist., iv, p. 87. \! Song of N. Necht (Catal. Ir. mss. Brit. Mus., p. 640, and Egerton, 138) of Namat, Keating, i, p. 219; Metr. Dind S., x, p. 27. Wusrropp—The Eurthworks, Sc¢., of S. E. Co. Limerick. 147 sent Hogabal and ‘Aine to Cliu. His wife, Boand, ‘of the silver fore-arm ” or “silver yoke,” was drowned by the hazel well, which similarly drowned Sinann, who was also Dairine. It was identified with the Severn,' beside which lay the shrine of Nodens; Cuil, another wife of Nechtan, was patroness of the chief cemetery of the Dergthene at ‘Oenach Culi (or Clovhair), near Knocklong.’ By this table it is evident that 1 to 5 are identical with the silver-handed god, ‘son of Hochaidh, father of Tadg,” and predecessor of Bress; that 7 and 11 are the same; that 4, 7, and 8, the father of Glas, the red, or red-handed, god, are identical, and 4 identifies them with the first group; that 4, 5, and 10, relations of Sithbaice and Dairfhine, are identical, and only the vague Nuada Fullon is unidentified save with Bress, the successor of Nuada Argetlamh. Nuada was a British god, not hitherto found in any Gaulish place-name or inscription.’ So far from being “an Irish high king,” he was (as we noted) a purely divine nature power, god of lght, of the deeps (of the earth and sea) and of war. Thanks to the discovery of his temple in 1805 at Lydney Park, Gloucestershire,' and his representation in metal there, we know how his worshippers regarded him. Tablets are dedicated to him as “ Devo M. Nodenti,’ or Nudenti, or Nodonti. He rides in a chariot drawn by four horses, wielding a club, his head decked by a spiked crown,* or by rays. The winds fly round him ; he is surrounded by tritons and sea monsters, and a fisherman catches a salmon." His temple has three cellae, so perhaps he was one of a triad, as at Sid Nechtain, on the Boyne. Many figures of wolves were found, and “ Nudd the superior wolf lord” figures so late as the sixth century in a poem of Taliessen. London and Ludgate were traditionally connected with his name “Ludd,” and he may have been the god whose temple lay where rises the great dome of St. Paul’s, Being a leading war god, he may be identical with Nedem (Net),’ Camulos, and Segomo. Net’s wife Nemetone (Nemon) 1* They, however, are not connected with the group of Clfu, and would overcrowd an already overcrowded paper. Bops Drrc.—Of all the gods of Munster “Bodbh Derg,” by a general consensus, was ruler.’ This probably sprang from his holy hill of Slievena- 1 Rey. Celt., xvi, p. 276. 2 Sanas Chormaie, p. 114. 3 «© Voyage of Bran,” ii, p. 213. 4 Rey. Celt., xxxv, p. 6-9; Encye. Relig., p. 6; Folk Lore, xxviii, p. 181. The Rennes Dind Senchas (Rev. Celt., xvi, p. 276) makes him a druid anda champion. In Achill he is ‘‘a king of druidism and enchantment and deyil’s craft.” (Larminie’s ‘*Trish Folk Tales,” p. 1.) 5 So the Gauls (like all nations) had their holy mountains where gods were called Paeninus (Pennine Alps), Vosegus (Vosges), and Dumiatus (Puy de Dome). See Nneye. Relig., viii, p. 863. ° Mesea Ulad, p. 15. 7 “Legend of Cacht,” Book of Lismore (Eriu, yili, p, 55). R,1.A. PROC., VOL. XXXIV., SECY. C, [23] 152 Proceedings of the Royal Irish Academy. man being in the chief and oldest territory of the conquering Dergthene, as their special goddess, Aibinn, became chief banshee of Munster in later days. He was at least affiliated to the great gods of the Boyne, being “son of Aengus of the Brugh, called the Dagda”; but this was a confusion, as the Dagda, “ Eochaid Ollathair, greyer than the grey mist,” was evidently his reputed grandfather. There seems reason to believe that Oengus was a pre-Celtic god, unknown to Gaul and Britain, whose vast chambered cairn of Newgrange, the noblest monument of the Bronze Age in Ireland, impressed the early conquerors with reverential awe. Invading polytheists were ready to accept and bring into their own pantheon any local god. Even Israel took over the Baals, along with the vineyards and tillage of their worshippers ; and the lion-worried settlers in Samaria, and the townsmen of Gezer, took over the worship of Jahveh. A local god “understood the country,” and was worth winning as a friend. As Bodb and the sea god divided the holy mounds among the gods,! each was evidently a “commissioner” for a different divine race. Others said that Bodb was “son of Eochaidh garbh, son of Duach temen, son of Bress, son of Elathan,? which has this much in its favour, that the predominance of Nuada in the Dergthene beliefs may have brought his successor, Bress, into prominence, and, as we know, Bress was celebrated in the legends of Cenn Febrat. A “Derg of Sid Derg” appears among the Tuatha Dé at Magh Tured, and, in other legends, the Dagda (not Bodb) allots the mounds, and was himself preceded in the ownership of the Brug mna Elemair by Elemair and his wife.‘ “ Bodb, over the S/d assemblies of Munster,” dwelt on the stately dome of Slievenaman, in southern Co. Tipperary, probably in the cairn on its summit.’ It has, however, a second cairn, an acknowledged S/d, on Knockshegowna, one of the spurs of the hill, which bears the same name as another famous fairy hill, farther north, in Ballingarry parish. The great cairn is about 86 paces round, and 9 feet high. Close to it (as at Uisnech) are remarkable natural rocks, one, with a slab resting upon it, and called “Finn’s table,” 5 feet 4 inches by 4 feet 6 inches by 3 feet thick; and, farther south, 1 Tr. MS. Series, R. 1. Acad., i, p. 46. We also find Bodbh Derg and the god Midir presiding at a council meeting of the Tuatha Dé, Silva Gadelica, 11, p. 116. 2 Silva Gad., ii, p. 522. 3 Also aided Conn at Magh Leana (‘‘ Magh L.,” p. 90) ; see also ‘‘ Cath Finntragha,” pp. 13, 239; ‘*Coir Anmann,” iii, pt. 2, p. 235. ‘ Dinds. of Carn Chonaill (Rev. Celt., xv, p. 478) ; Book of Fermoy (Ir. Texts, R. I. Acad., i, p. 46) ; Encyc. Relig., iii, p. 283. 5 On the Nilgiri Hills in Southern India each deity was associated with a peak on which was a stone circle, barrow, or cairn. Wesrrope—The Larthworks, §c., of S. Lb. Co. Limerick. 158 resting on four pillars, is another stone, 7 feet by 4 feet.' | These probably mark Bodb’s holy place. Nuada Derg had a fort on the same mountain, and may, perhaps, have been identified with Bodb Derg* if the Dergthene tribe took over the worship of the god of Slievenaman. ‘he hill, as we saw,’ forms a conspicuous and beautiful landmark, when looking down the vale of Aherloe, from the hill of the goddess Aife, over Gleneefy and Duntrileague. Bodb seems to have had two residences—the Sid, so widely celebrated as Sid na mbann, Sid bann finn os Femen, and Sid Femen; and a second, some- where near Loch Derg. Some said he was brother of the goddess Dechtire, and uncle to Cu Chulamd. He was a god of the move amiable type. When pressed to let his attendants pursue and wound Ler, the sea god, he, with calm dignity, forbade it, saying, “I am none the less king of the Tuatha Dé Danann because he is not submissive.” He also controlled his anger when the wicked Aoife said that her husband Ler would not trust his children with him. Like Lug, he is called “one to whom all science has done homage.” He had friendly relations with the mortal rulers of the land, and made a treaty with King Conn of the Hundred Battles, undertaking that neither he nor the Tuatha Dé under his rule would injure the king. Fer Fi of Aine was a hostage, or perhaps a guarantor, of the treaty. From his taking the place of Nuada Derg in some versions of the Mog Nuadat legend, I venture to suggest that the Dal Cais endeavoured to identify the chief god of their new settlement with the more familiar god, Nuada, whom their ancestors had worshipped before their invasion of Ireland. Bodb had a numerous offspring.’ His “seven sons” and his “ three sons” are named. Among his children are Ferdoman; Artrach, who had a brwiden of seven doors, and a diss, Rath Artrach, once resplendent, but blighted even in Finn’s day. Bodb’s daughter, Sadb, was plighted to Finn at Sid Femen ;’ but (of course in so evidently late a poem) she was mortal, and her death was announced to “her highborn kinsmen” of the Sid, the Tuatha Dé. Bodhb’s three sons were born in his father’s Sid at the “many-windowed Brugh of the Boyne.”* 1 Ordnance Survey Letters, Co. Tipperary (R. I. Acad., 14, F. 18), i, p. 170. One recalls the altar of Zeus on Mount Lycaon, a mound with two pillars facing the sunrise, as described by Pausanias. +“ Magh Leana,” p. 3. 3 Supra, Xxxiii, p. 479. * Book of Leinster, p. 123b. 5 “Wate of Children of Lir” (Atlantis, iv, pp. 115-131). § Book of Fermoy (Ir. Texts, R. I. Acad., i). ‘The poems connecting Finn with Slievenaman are endless, but more striking is Keating's note of Suidhe Finn on that mountain (History, ii, p. 325). S Agallamh, ii, pp. 106, 171, 225. 154 Proceedings of the Royal Irish Academy. Bodb’s dependents were less estimable. He kept a sort of training college, which gave “ S/d na mbann” its name, “the mound of the women.” Legend speaks unfavourably of the management. Bodb’s own daughter was abducted; so was Deltbanna, Conaire’s attendant, the wife of Life and daughter of Cannan;' and Finnchada, who was carried off from it by Sengarman, so late as the reign of Dathi, in the early fifth century. Some said that Connla, son of the High King, Conn, when he was lured away by the lovely woman of the S/d, went with her, not beyond, or beneath, the sea, © but to Sid Bodaige, by which Sid Buidb is probably intended.* Cliu tried to lure Conchenn, another daughter of Bodb, out of it. In it, too, was edu- . cated Aoife, the cruel wife of Ler, who turned his children into swans. Another scandal was attached to Sid Fer Femen, if it be the same place, © or the second cairn, when, during the absence of Oengus, son of the Dagda, on Samhain night, the god Midir carried off to it the unfaithful wife of Oengis, Englec, daughter of Elemar. Len Linfiaclach, a retainer of Bodb, of Sid Femen, gave his name to Loch Léin, or Killarney Lake, in Luachair Deadaidh.t His servants were also of doubtful character. His swineherd, “* Nar-squinting-with-the-left-eye,” sits cooking squealing heads of pigs over the fire, aud bringing bloodshed to every feast which he “graced” with his presence.”* Two other swineherds, Friuch and Rucht, were turned to ravens.° Yet another, Tummue (in the form of a many-coloured beast), was swallowed, and became the Donn Bull of Cualnge, as he revealed to King Ailill. It seems doubtful whether the second residence, S/d Buidb, was near Portumna, or near Killaloe. It was beside Loch Derg. Perhaps it was the conjoined rings above Derry Uastle, or even the Sid of Craglea; but 1 have no evidence. Tra Buidb Derg, on Achill, also bears his name. I know nothing but the mere dictum of O’Curry’ to regard Bodb Derg, king of Magh Femen, and his namesake, the son of the Dagda, of Loch Derg, as different persons. O’Curry had a childlike faith in the historic character of the Tuatha Dé. In the “ Dream of Oengus” the king of Magh Femen evidently helps Oengus on his own ground, when he wins him the swan maiden on the lake in the Galtees. "** Magh Leana,” p. 1 (and ‘* Tochmarch Momera,” p. 155); Rennes Dind S., Rev. Celt., xv, pp. 303, 318 ; Atlantis, iv, p. 119. I do not know whether it and the *‘ Sid of the men of Femen ™ be the same, or whether they be the two cairns of Slievenaman and Knocksheegowna. * Silva Gad., i, p. 534. * Metr. Dind S., x, pp. 41-43. ‘ Rev. Celt., x, p. 446. Noisiu, son of Nechtan Findguala, was also connected with the lake. (Jbid., xvi, p. 79 and p. 451.) ® Rey. Celt., xxii, p. 313. * “ Voyage of Bran,” ii, p. 58; Folk Lore, xvii, p. 166. * Atlantis, ivy, pp. 156, 169n. Wesrropep—TVhe Earthworks, §e., of S. B. Co. Limerick. 154 Bodb is nearly forgotten, but, like most local “ fairies,” clung to his own home, and won some little meed of recollection down to the last century. CairBrE Musc.—The Museraige tribe bears all the mark of having been broken up by a series of invasions; “ islands” of it persisted through Munster, Musceraighe Liac Thuill, or O’Noonan' of the Ui Fidgeinte in south-west Co. Limerick; Muscraige Mitaine, in Cliu among the Mairtene, in the Galtees; in Muscraighe Treitherne, Muscraige Chuire, Museraige Tire, and Museraige Breogain, in Tipperary, Musecraige Luachra, still Muskerry, n Cork, and small fragments at ‘Tontinna on Lough Derg, in east Co. Limerick, in Magh Femen, and elsewhere.” There was some close bond between it and the Dergthene, especially with the Dal Cais. In legend it was closely connected by descent with the Corca Duibne in Kerry and the Corea Baiscinn in Co. Clare. When the tribal genealogists attempted to “ regularize ’ the position of non-Milesian tribes (long before the euhemerist movement), they divided the family god, or eponymus, into a triad, the eldest of whom was (Oengus) Cairbre Muse. But so far from the group belonging to the third century, the Tain bo Cualnge mentions “the three Cairbres from Clin” as warriors of fame. More than two centuries earlier legend made them sons of Conaire, High King of Ireland, successor of Conn of the Hundred Battles, “B.c. 82.” His legend‘ was certainly pagan, and outraged all Christian ethics. His epithet Muse was wrenched into Mo Aisye, “inordinate desire,” for the Muscraighe sprang from him and his own daughter, Duben, or Dovin, whence Corea Duibne. The territory covered the modern Corcaguiny and Iveragh to Valencia Island (Dairbre). In the former section we find several ogmie inscriptions with her name, “ Maqi mucoi Dovinnias,” one on a noble site, on the entrenched headland of Dunmore, overlooking the Blaskets. It seems very clear that mythology ran mad on this legend. Duben was possibly (as seems implied) the mother and then the daughter of Cairbre ; ? then, to get out of the vicious circle, she was turned into ason. Cairbre’s name was equated with Corbad, parricide, by some local Fluellyn. His name was said to have been Angus, and he and his brother slew their stepfather, Vlizabethan Surveys. For the race of Conaire mér and the coming of the Muscraighe from Magh Bregh see Eriu, vi, p. 138, from Books of Lecan and Ballymote. Muscraige tri muige is the Aes tri maige in N.E, Co. Limerick, * See notes on O’Huidhrin’s ‘‘ Topographical Poem ” and the Onomasticon Goedelicum. 8 Loe. cit., p. 351. ' Studied in R. Soc. Antt. Ir., xl, pp. 184-5. > Keating, Hist., ii, sect. xli, p. 278, tells another story, where Cairbre Musc slays his stepfather Neimidh in his mother Saruit’s arms—another proof of the varying nature of these tales. He derives the epithet Wo Aisge from this event. See also Eriu, vi, p- 143. 156 Proceedings of the Royal Irish Academy. husband, Oilioll Aulom, and Fiacha, his favourite grandson, gave Cairbre Musc all eastern Co. Limerick from Aine to Lough Derg. The mythology of Cairbre Muse seems to have no foothold in fact, and abounds in contradictions. On the birth of his misbegotten sons all nature sickened, and the crops failed. The tribes ordered that the children should be burned, but a druid saved one, Core, putting it on a red cow’s back. After a year the sin entered into the cow, who swam out to sea, and became the “ Cow Rock,’ Bo Bui Rhys,* in a very elaborate study (which students should master and assess for themselves), equates Cairbre cinn chait, Cairbre, father of Ere ; Cairbre, the enemy of Bress; Cairbre Musc, the owner of the dog, “Mug Eime,” and MacKinealy (the slayer of Balor, in the Torry Island legend), with the Gaulish “Mercury” and Lug. Be this as it may, Cairbre Musc is evidently from first to last a mythical personage and a tribal god of any date before the time of the Raid of Cualnge to the third century. We are confronted with a difficulty. Keating? alleges that the descendants of Conaire alone were the Ernai. He tells how Mog Nuadat expelled the latter from Munster, ““so many as would not submit.” Conaire had an ancestor, Daire Doinmhar, who was possibly confused with “ Daire ”’ or “ Dairine ’’ in the Clann Dedad and Corca Laéghde pedigree. Conaire may be confused with another High King, Conaire Mor (B.c. 110, son of Eidersceol, a descendant of Degad, son of Sin, and of Oilioll Erann, and suecessor of Nuada Necht), from whom the Clann Dedat claimed descent. The three tribes possibly were branches of the Eynai by descent or affiliation. Indeed, the louernoi of Ptolemy seem to overspread all south- west Ireland up to the Vellabori. The latter have of late been placed as far north as Tralee, but their position may be anywhere from Iveragh to the “ Dour.” Cuiv.—Another eponymus of the extensive district of Cliu mhail mhic Ugaine, in south-east Co. Limerick and along the Galtees to the Suir, meets us in our district. The mighty mountains bore his name, Crotta Cliach, “ the harps of Cliu,” which I venture to suggest are the two harp-like cooms, with strings and frames of stream gullies, seen on the flank above Aherloe. Cot, however, is probably a pre-Milesian word, as a tribe named Crotraige dwelt along the foot-hills, and the name possibly gave rise to the folk-tale. Cliu may have been connected with the goddess Clidna, for the Mairtene, another non-Milesian tribe, near the Galtees, were called “the champions of Clfu and the contentious hosts of Cliodna,”* suggesting such terms as “champions of 1“ Leabhar na hUidhre,” 54a, and ‘‘ Book of Leinster,” ‘‘ Magh Leana,’’ p. 28. The name Bui is found in the Portolan Maps, 1450 to 1593 (Proc. R. I. Acad., xxx, p. 417). 2 Hib. Lect., iv, pp. 317-321. 3 Hist., Book i, sect. xl (Ir. Texts Soc., viii, pp- 231, 263, 269). ‘ “ Magh Leana,” pp. 77, 78; the passage in our present copy possibly refers to the Westropp—Vhe Earthworks, &c., of S. EB. Co. Limerick. 157 the S¢d,” “ champion of the cairn,” “champion of Segomo,” and “ champion of (the goddess) Nav.” Legend tells how Cliu was a harper who came to Baine’s Sé, a place so sacred that the very “ Book of Rights’! forbade anyone to walk on it “ by the light of red fire,” like the mounds of Shevereagh, that men dared not sit 4 999 upon “for dread of the Tuatha Dé,” and the procession with flaming “ wisps ” at Aine’s mound. Cliu became harper to King “ Sinirdub,” and used to play on two harps at once, whence the mountains were called Crotta Cliach. He went to the “ Sid of the Men of Femen,” where Bodb dwelt, to carry off his daughter Conchenn. The inhabitants tried to drive him off by magic, but in vain; so, after a year, the lady Baine lost patience, and sprang out in dragon’s shape, when Cliu died of the fright. Hence the lake was called “Toe bel draceon.’* It was evidently not the place of its name in Westmeath, but the lake in the Galtees, at which Oengus of the Brug, aided by Bodb Derg, won his swan-wife, “Caerib Ormeith,” daughter of Ethal Anbual of Sid Uamain in Connacht, whom he found with her transformed maidens sporting on Loc bel draccon.* The Leabar Breac* tells a nearly identical story, but the king is called “ Smirdubh mac Smail” of the Three Rosses of Shab Ban, and Cliu seeks Bodb’s daughter at Sid Femen, till the water bursts up under his feet, and forms Loch Béal Sead, on the mountain. On this ‘‘ Coerabarboeth,” swan-daughter of Ethal, and her fifty companions float ; it is also “ Loe Crotta Clach” and “ Loc bel dragan.” It is evidently some lake, still unidentified, on the Galtees. Rarely in Western Europe do we tread so closely in “ the footsteps of the dead old gods” as around Cenn Febrat and the Galtees. Curt anp Necutan.—Nechtan and Nuada Necht are sometimes taken as son and father, and sometimes as identical. Nechtan encouraged the family of Eogabal to settle on Knockainey ; and Cwul, wife of Nechtan, was patroness of the chief cemetery and fair of the Dergthene, called ‘Oenach Chulimna, Nechtain, and ‘Oenach Clochair." Nechtan brought the first pregnant cow to districts, the word Nia not ocecurring—yet there is no evidence for a district of Cliodhna near Kmly in the Mairtene territory. 1 Book of Rights,” pp. 5, 21, the present Knockmany, a Bronze Age cairn chamber. There was also a Cnoc Maine near Kilfinnan (Silva Gadelica, ii, p. 123), probably the place referred to as Sid Bain in the Cliu legend. Could it be the mound of Kilfinnan ? 2 Agallamh, ii, p. 124. 3 Rennes Dind S. (Rey. Celt., xv, p. 441). 4“ Dream of Oengus”’ (Rev. Celt., iii, pp. 847-355). 5 “ Manners and Customs,” p. 246. 6 The place-name ‘‘Sciath Nechtain,” where Olchobar, king of Cashel, defeated the Norse in 847, suggests the disc barrow of ‘‘Cu Chulaind’s shield” at Tara and ‘*Sciath Gabhra’’ where the MacGuires were inaugurated. There was a ‘‘Sciath na bfeart” (shield of the burial-place) in Roscommon, and another ‘‘Sciath Nechtain” in the RKoghanacht territory. 158 . Proceedings of the Royal Irish Academy. Treland,' and the euhemerist Flann (in 1050) says he was killed by Sigmall, grandson of the god Midir ;* but this is of the warped stratum of legend, not found in the early myths. Some (by the q to p change) identify him with Neptune, but this is more than doubtful = and unless he is Nuada (with whom even then we have to go back to Nodens. in Britain, to find him a sea- god), he has no connexion with the ocean. Old writers equated echt with Sneacht, snow, because “ Nuada Necht was as white as snow,”’a folk derivative. He was certainly Nuada, ‘fair head” or “red hand,” a king of Munster, whose rival, Bress, was slain by Lng’s device at Carn Ui Neid,* for Nuada Argetlammh is superseded by Bress, and the king of Munster seems to be Nechtan. Nuada, his wife Boand, and son Nechtan, all abound in epithets like “ white” and “ silver.” Nechtan reckoned the Corca Laegde amongst his human descendants. They owned a fort, Dun mic Nechtain, near Kenmare. I fancy the Co. Clare fort, Cahermacnaughten, is named from a human family. Near Kenmare ‘“‘ River,” Inber Scene, dwelt also a namesake, a lady, Necht. He and his three cup-bearers could alone approach to look on the sacred well at the source of the Boyne, and a Loch Nechtain (so far unidentified) lay among the Echtge Hills.’ Despite her evident importance, I have so far found no legend of Cuil. She may lurk under an alias name, for what was told of the Hazel Well and the fate of Nechtan’s wife Boand was also told of Sinann, Macha, or Dairfhine, the Shannon goddess, who had many alternative names.* As regards Sinann, her sister “‘ Echtach” is often confused with “Echtge the Horrible,” the goddess of the mountains of Echtge, who was reared on children’s flesh, and was given the mountains by her lover, Fergus Lusca Beist. The latter was cup-bearer to Gann and Genann,’ who, with Sengann, ' Egerton mss., 138; Catal Irish uss., Brit. Mus., p. 640; R.I. Acad. ms., 23. D. 13, p-. 146. = Leab. Gabh., p. 183, ‘‘on the Deaths of Tuatha Dé."’ * Dr. O. Shrade, ‘‘ Prehist. Antiqq. of the Aryen peoples,” ed. F. Jevons, p. 412; Hib. Lect., iv, p. 123. ‘ Irische Texte, iii, 368; Folk Lore, xvii, p. 30. * Rennes Dind S. (Rev. Celt, xv, p. 439); Metr. Dind S., x, p. 216. ® Rennes Dind S. (Rev. Celt., xvi, p. 83). * The name Nechtan is found among the Ui Fidgeinti in Co. Limerick in historic times. - * Boand (it should be noted) came from the South to Meath (Metr. Dind S., X, p. 37. “‘Boand I’). Nechtan Nuada is son of Labraid, whereas (in Boand II) he is son of Namat ; see also Rev. Celt., xv, p. 293, p. 430. ® Rennes Dind S. (Rev. Celt., xv, p. 458) ; Metr. Dind S., x, pp. 305-7. In us. R. I. Acad. 23 D 16, p. 411, is an amazing forgery, an Irish inscription found on a coffin with a skeleton 12 feet long of Genann, king of Ireland, at Breaffy, Co. Mayo, in 1732. This was written in 1756, and adds that Genann had oxen 9 feet high! His brothers Gann and Sengann landed at the Inbir Dubhglais near Lehinch, in Co. Clare (Eriu, viii, p- 13). Wesrropp—The Earthworks, §c., of S. BE. Co. Limerick. 159 were eponymi of the tribes in Co, Clare and Connacht, called Ganganoi by Ptolemy. She was of the Tuatha Dé, and reared at Cuil Echtair, near S/d Nennta.' Altogether, the important position taken by the Irish goddess corresponds to the high position of the mortal women, queens, and warriors, like Boudicea and Medb. It is notable that great sanctuaries, like Knockainey and Oenach Chuli, are called after women, and that Carman, Tea, Macha, Tailti, and Maistiu loom so large at Carman, Tara, Telltown, Emania, and Mullaghmast. Human sacrifice was offered at Emania, Tara, and ailti, perhaps to the mother goddesses of the harvest. DEDA, GARBAN, AND FEBRA.—The great plateau of Cen Febrat, or Slieve- reagh, so prominent an object over all eastern Limerick, was, as we saw, the chief cemetery of the Erann, or Ernai, the Clann Deda, and Corca Laegde. All the outstanding names of their mythic ancestry, save Deda? and Curoi (the latter presumably buried at Caherconrigh), are attached to its monuments and mark it as the great cemetery of the Ernai at Temair Erann,* so wrongly placed by O'Donovan as near Castle Island. In the later frantic attempt to unite all the tribal genealogies, the Ernai and Muscraige are given “ Ailill Erann, the god of the bolg ga,”* as common ancestor; but. in most docu- ments, the Clann Dedad Ernai derive from Deda, son of Sen. Sen® was probably father of a large group of gods of Mountains and Lakes; Deda gave his name to the Clann Dedad, or Degaid, and to Luachair Dedad.*° Sen was son of the High King, Eochu Airiomh, “ who was seventeenth in descent from Ugaine Mor, and first dug a cave in Erin,” and reigned from “B.c. 125 to 110,” according to the chronologers. Deda and his brethren of the Clan Rudhraighe were expelled by the line of Eremon from Ulster. They fled (in one story) to Duach, King of Munster, who gladly received so valuable an army and gave them lands, so they became the Ernai (Ptolemy’s Iouernoi). When Duach died, the High King Eochu Feidlioch (Queen Medb’s father), 1 Some made her daughter of Deda, son of Sen (see infra). 2 Are the Clanna Deagaidh of Deda’s line the Clanna mac Deichead or Maqi Deceddas of the ogmic inscriptions? (R. Soc. Antt. Iv., xxxii, p. 30, and R. I. Acad. Proe., xxvii, p- 339.) 3 Tract on the cemeteries. Cf. Metr. Dind S., x, p. 227. See section at end of this paper. 4 New Iv. Reyv., xxvi, p. 133. He appears in three places in the tribal pedigree—at the beginning, in the middle, and some generations after Deda—another proof of the uncritical usage of old material by the genealogists. Keating makes him 9th in descent from Ugaine and 8th in ascent from Sen. 5 Sen was in some documents one of a triad Deitsin, Sin and Roisin, e.g., Keating ii, p- 289. 6 Vol. ii, pp. 229, 235, 237. 7 Keating, ii, pp. 158, 184, 229. Todd Lect. Ser., iii, p. 197 ; Metr. Dinds., x, p. 241. Keating ii, p. 158, p. 229; Ossianic Soc., v, p. 288n. R.I.A. PROG., VOL. XXXIV, SECT, C. [24] 160 Proceedings of the Royal Irish Academy. B.C. 137, confirmed Dega as King of Munster, to which he had already been chosen for his prowess ;' but others said he was confirmed by Lugaid Luaigne, grandson of Nia Segamain,? so he became an eponymus of the Clanna- or Corca-Deagaid,® “many a spear was in its host in the time of Deda, son of Sen.” So far the story is moderate and even probable, especially if the invasion of the Mile tribes took place not long before our era, as some have supposed. The Ernai Dedad, with the Ultonians and the Feini of Tara, were named as “the three noble races” of Ireland in the Senchas Mor. Dega had forty sons; chief of these was—(1) Daire,’ father of the famous Curoi, King of Munster, in the Red Branch* tales. Curoi was rival of Cu Chulaind, by whom, and by the treachery of his wife, Blaithnaid, he fell ;7 and his son Lugaid avenged him by slaying the “ Hound.” Forgoll, another son of Daire, was ancestor of Finn and Oisin; and Lugaid, a third son, had a wife Eithne, said to be buried in the group of dise barrows at Cooloughtragh in Temair Erann. (2) Taeth, from whom Cain, son of Fergus “ Fer Deda,” was named “ Mogh Taeth”; probably both names are dedicatory to the god Deda and his son, (3) Iar, father of Eterscel, father of Conaire. (4) Fir Ceit, ancestor of the Dal Ceiti, in Munster. (5) Cleite, who died at Cletty, on the Boyne, named after him. (6) Dea, from whom Inbir Dea, the Vartry mouth, was called. (7) Garban, or Garman, from whose grave burst Loch Garmain, or Wexford Harbour. (8) Senach, who gave Sliab Mis (Mish) to his bride, Mis, daughter of Cairid. He died, and was buried in his rath there. (9) Ross, the famous Red Branch warrior; he took part in the overthrow of the sons of Umoir. (10) Coemgin Hornskin, in some versions, is said to have married Mis, and given the mountain to his brother, Senach the Rough. (11) Conganchness, perhaps the same as the last, avenged his “ brother” (nephew) Curoi, and devastated Ulster till Celtchair of Downpatrick slew him with red-hot spits. (12) Cabalglinni, or Faelglinni, the senior of the tribe in the Mesea Ulad. He also had a daughter, a mountain goddess, '* Cathreim Conghal Clairingneach ’’ (Ir. Texts Soc.), pp. 2. 3. ? Rawl ms., 502, p. 1488. > Rev. Celt, xx, p. 336. ‘ Metr. Dind S., x, p. 236. ® Usually identified with MacNiadh’s grandfather Daire (Keating, iv, p. 115, Genealogy of O'Driscoll and Miscellany of the Celtic Society, ‘‘ Corca Laidhe,”’ pp. 25, 9, &c.) ® Loc. cit., p. 80. - * Rennes Dind S. (Rey. Celt., xv, p. 448) and much other material. For the other sons (2) Coir Anm., p. 407 ; (3) Metr. Dind., x, p. 241; (4) Proc. R.I.A., xxix, pp. 81-87 ; (5) Book of Leinster, f. 166b; (6) Rev. Celt., xiv, p. 429; (7) Rev. Celt., xv, p. 428 ; (8) Rey. Celt., xv, p. 445; (9) Zhid., p. 478; (10) ‘‘ Death Tales,” Todd Lect. Ser., vol. xiv, p. 27; (11) Metr. Dind S., x, p. 241 ; (12) Mesca Ulad, p. 41. Wesrropp—The Earthworks, §c., of S. E. Co. Limerick. 161 Echtghe the horrible! There are a number of variant tales*—the Coir Anmann® tells how Duach Dallta Degad, son of Cairbre Luse (the lame) and ancestor of the Dergthene, had two sons, Duach and Degad (B.c. 168). The latter, being the popular candidate for the kingship, was blinded by Duach. Another story tells how Duach fell by the Ulstermen and Degad, son of Sen, King of the Ernai. The hero Finn descended‘ from Daire, son of Deda; while his mother, Murni, was derived from Tadg, son of Nuada, son of Achi. Clann Deda held all Luachair® to Curoi Mac Daire’s fort on Cen Febrat. Though Deda is nowhere (to my knowledge) called a god, he seems ancestor and father-in-law of a number of eponymi, mountain gods, and connexions of gods. His fame grew on no mortal soil, but is all mythic and superhuman; we need hardly look for a nucleus of fact, if such ever existed. Deda’s issue, Corco Deda and Corca Muige (both in western Co. Limerick), are in the list of the Aithech Tuatha.’ Probably the pre-Milesian tribe had its own divine beings and ancestry which later scribes tried to euhemerize and connect with those of other tribes. Febra, son of Sen, Deda’s brother, was slain by Cain Dercedualach, husband of the goddess Aife, sister of Aine and Fer Fi; from these heroes the mountain derived two of its names, Cenn Febrat and Sliabh Cain. _ Garban, son of Dedad, slew his uncle’s murderer, and he and Febra, when the gods became mortal, were reputed to lie under two of the pre-historic mounds in Temair Erann cemetery, probably those to the south-west of the spring at the gorge of Glounacroghera.’ South from them, Iugaid Laigde, the eponymus of the Corea Laigde, was buried. Garban’s other grave, the source of Wexford estuary, we have already noted; he is called ** Garman Glas Mac Degann,” so his identity is well established. Eithne and three other ladies lay in the conjoined rings of Couloughtragh. She was niece of Curoi. Ere of Inluacra probably lay in the mound in Ballinvreena.* This 1 Metr. Dind S., x, p. 305. The ‘‘ Cath Ruis na Rig” (Todd Lect., iv, p. 25) names a contemporary of Curoi, Eochu, son of Luchta, king of North Munster, and the Recar- taig Dedad, at Temair Luchra. *The genealogists ring many changes on Sen, Deda, Daire, MacNiadh, Lugaid, MacCon, and Lugaid Laegde, and move their period over a space of thirteen or fourteen centuries in some instances. 3 Trische Texte Ser. iii, pt. ii, p. 292. +“ Puanaire Fionn” (Ir. Texts Soc.), introd.; Yellow Book of Lecan, col. 768; Ossianic Soce., iv, p. 284; Silva Gad., ii, p. 245. 5 Mesca Ulad, pp. 17-19. 5 Rey. Celt., xx, p. 336. * Metr. Dind S., x, p. 231, p. 247 ; Rennes Dind S., Rey. Celt., xv, pp. 115, 442 ; supra, XXxilil, pp. 460-466. 5 Possibly Erc, son of Feidhlimidh, drowned in a lake (Miscellany, p. 61). [24+] 162 Proceedings of the Royal Irish Academy. . section gives a faint impression of what a rich mass of legend must have existed about Luachair and its “altered gods.” Now I have used the term “Ernai,” not only for the Corea Laegde, but also for the Clann Degaid, and Keating (ii, p. 313) says that it was wrongly applied to these, and should be only used for the descendants of Conaire Mor, the “ Museraige,” Corca Duibne, and Corca Baiscoinn. This seems doubtful ; it is evident that the Clann Dedad and Corca Laegde had the same cemetery and sanctuary on Cenn Febrat, and oceupy the very position where the *‘Touernoi” are shown by Ptolemy. The Corea Oiche, possibly, are also of this kindred, though they, like the Ui Fidgeinte and the Corca Laegde, affiliated for political reasons in later years to the Dergthene tribes. The Sanas Chormaic (p. 16) speaks of the Corca Laegde or “Dairfhine” as descendants of “ Daire Doimthech,” and it seems evident that an early pedigree existed in which he appears with certain gods and ancestors—Lug, son of Ethleann, Deaga, Sithbolg, Nuada Neacht, and a group consisting of “Daire Sirchreachtech, Lugaid Laide, Lugaid mac Con, Mac Niad, and Duach,” and a Sen Lugaid;' Dairine was equated with Daire. The pedigree was old enough to find a place in the Saltair of Cashel. When the Corea Laegde tried to draw nearer to the Clann Deagaid, they naturally identified the pedigree names with the latter’s ancestors, Sen, Deaga (Deda), and Daire.’ Lugaid Laegde was placed sometimes as fifteenth in descent from Ith (circa B.c. 900), sometimes a contemporary of Eogan and Art (about a.p. 300); Daire was first called “King of the West,” then “ King of the World,” and reached the zenith of his power in the Cath Finntraga, when he invaded Corcaduibne. The Book of Lecan made him father of Mac Con and the three Fothads, mythical kings of Ireland (A.p. 296). Finally, it is evident that, when the Dergthene tribes of Cashel and Thomond became powerful, the Corea Laegde inserted “ Dergthene” into their own pedigree as a common ancestor, as the Tradraige, after attempting apparently to affiliate to the Caenraige and Corca Baiscoinn, asserted their kinship to the Eoganacht of Cashel to win favour of the powerful King Fedlimid,*? about a.p, 840, When we recall the extremely mythical legend of the three Cairbres, we see that we are not to let such tales weigh against the general fact that, like Crim- thann, Nia Nair “of the Ernai” at the beginning of our era (one of whose forts stood on Slievereagh, near the Ernean cemetery), the term “ Ernai” ‘ Identified by some with Lughaidh Laeghde ‘‘ Magh Leana,” p. 9. *** Miscellany ” of Celtic Soc. (‘*Corca Laidhe”’), pp. 4, 5, 9, 57,59. Cf. pedigrees of Duald MacFirbis, Roderic O'Flaherty, in ‘* Ogygia,” and Keating, iy. *R. Soc. Antt. Ir., xxxvii, p. 407. The Corca Oiche and Corca Muicheat are of the Atheach Tuatha at first, but are affiliated to the Dal Cass after 890. Wusrropp—The Lurthworks, §¢., of S. L. Co, Limerick. 168 may be applied, like the term “ Fir Bolg” or “ Milesian,” without any allega- tion as to blood kindred in the races so described. In Daire, Ded, Mis, and probably Lugaid Laegde we are dealing with gods, or demigod ancestors, not with facts. DerGruENe.—We find in the Ancient Laws! that “deirbhfine and deirgh- fine” were tribal divisions, so one suspects that the ancestors from whom the Dergthene and Dairfhine are named were mythical and not even personal.’ Perhaps the descendants of Nia Segamain, in the flush of their invasion of Magh Femen, called themselves “ Ze Tribe.” In some of their pedigrees, indeed, a triad “Dergfhota, Deirgthenic, and Deirg”’ take the place of the single ancestor, the father of Mog Neid. The Corca Laegde pedigree makes him son of Nuada and great-grandson of “ Lug mac Kthleann,” a god con- fessed. “ Dairine” may also have been a goddess, as the name is identified with Macha‘ and Sinann. We find a Core, son of Dairine, a king of the Corea Duibne, who accompanied St. Patrick to Ulster in late story.* Dairine was also father of Nuada “ Dearg” or “Salfota,” foster-father of Dergthene’s grandson, Mogh Nuadat,’ and has been identified with Daire, son of Sen. Lugaid mac Con, who is seventh in descent from Dergthene, is contemporary with Oilioll Aulom, fourth in descent from him—the name Lugaid being as common in the one pedigree as the epithet Dearg in the other. Nothing can show more clearly that these and similar tribal pedigrees are built out of disconnected fragments, not consistent or truthful, but real, archaic tradition, warped, but not pure fiction, and so worthy of critical study. - Donn.—This god has the merit of being free from the family complica- tions of the divine fathers of tribes. Donn Firinné’ and Aine are the two ancient deities who hold their own in Co. Limerick as Donn Dumach and Aibinn do in Co, Clare. The word Donn* has yet to be studied. It occurs in many a guise among the deities of Gaul, Britain, and Ireland, and is not merely the name of men-like gods (of both sexes), but of the Donn Bull, in ! Ancient Laws of Ireland, iv, pp. 282-5. * Sanas Chormaic¢, p. 55. 3 Book of Lecan, f. 215, also ‘‘ Miscellany” (‘‘Corca Laidhe’’) as Dergthenedh, Deagha Dearg, and Deadh mannra, son of Sithbholg, p. 57. ‘+ Thus Duben, the sister-wife of Cairbre Musc, was made his son, and Lugh’s mother, Hithliu, became his wife, Hithniu. Dairine was Macha, Nith, Neman, and Badbh (Rey. Celt., xxii, p. 58); Mo Febhis, whose son is Mog Ruith, is mother in L. na h-Uidhre, 74, and father in Ann. Four Masters, a.m. 3751. Hib. Lect., p. 526. 5 Agallamh, ii, p. 198. 5 “*Magh Leana,” pp. xxi. 2, 117. ? Firinné, the name of a Spanish Druid (ibid., p. 165). ‘ “Social History of Ancient Ireland,” i, p. 262. Dr. Joyce regards him as a son of Mil. The name ‘‘ Donnus”’ is found in a Gaulish inscription at Nimes (Rey. Celt., xiii, p. 303). 164 Proceedings of the Royal Irish Academy. Treland, and its counterpart, the Donnotaurusin Gaui! The Bull of Cualnge and his rival were reincarnations of human beings, and the step “from his brother the beast to his brother the god” was a short one in the ancient faiths. There are a great number of supernatural persons named Donn in later Irish literature; perhaps local forms of the same god. In the “ Battle of Ventry’”* we havea list of the (human) Tuatha Dé, who hasten to defend Ireland against the allied armies of the whole outer world. Seven bear the name, Donn from S/d bee Uisge, Donn Fritgrinne (? Firinne), Donn Teimneach Donn Senchnuic, Donn Chnuic an dos (evidently two hill gods), Donn of the Sandhill (Dumach, now Dough, on Liscannor Bay, Co. Clare), and Donn of the Swamps. In the “'l'riumphs of Congal Clairingneach,”* appears Donn, a Séd king, son of Iomchad, descendant “of the Dagda, of the prime stock of the Tuatha Dé Danann.”’ Another Donn, son of Fionnlaoch,' brought 100 women from the S/d of Aed (Assaroe, Donegal), to the Sid of Mis (on Slievemish, in Kerry). Among them was Aedh’s wife ; and she, jealous of her rivals, turned them into deer, and Donn into a stag, which was eventually slain by Bran and Finn’s other hounds at “Cenn Maghair,” on the coast. The Bruden da Derga* has a Donn ‘letscorach of the Séd mounds; the Agallamh has Donn, son of Midir, whose Sfd7 was assaulted once a year by the Tuatha Dé.’ Irish gods (like those of Greece) did not hesitate to fight their own kindred, or we might suppose him of some other race. We are also told that Lir of Sid Fionnachaid, “ the man who excelled in prowess all the Tuatha Dé,” was slain by Caeilte (a late mark), who also slew Lir, Dub and Donn of the forces of the Sid.’ It is impossible to regard the Welsh goddess Don (the Irish Donn) as in any way connected. Cassiopeia’s chair was her fort, Llys Dén, but at least we have the name Donn for a god, a goddess, and a holy bull. Donn, son of Midir, or ‘“‘ Donn of Uisnech,” was closely connected with our district, for his sons, Eogabal and Uainide, and the children of the first, Aine and Ferfi, loomed large among its gods. He was in touch with Nechtan, who recommended the migration, and therefore with Cail of Oenach Chuli. 1 think it very probable that the Donn of Firinne, the god of the long ridge, Hib. Lect., iv, p. 92n, Holder, i, c. 1307, see also ‘‘ Donnus,”’ ibid., ‘‘ Regius Donni filius.” * Cath. Finntraga (ed. Meyer), pp. 15, 260. 3 Ir. Texts Soc. (ed. P. MacSweeney), p. 77. * “ Dunnaire Finn ”’ (ed. J. MacNeill), p.131. Cf. other fairy deer, Metr. Dind S. vu, p- 11, and Rey. Celt., xv, p. 273, with a Christian equivalent in the Tripartite Life of St. Patrick. > Rey. Celt., xxii, p. 39. ® Agallamh il, p. 224 and (ed. Stokes) p. 140. 7 Agallamh, ui, p. 146. Westropp—The Earthworks, &c., of S. EB. Co. Limerick. 165 with its prominent cairn-topped dome, so conspicuously seen from ‘Aine’s cairn, was the son of Mider, and gave his epithet to it—Knockfirina. T heard as a child, about 1872, from the peasantry at Attyflin much about King Donn. His blue dome was an infallible weather-glass, whence its name, “The Hill of Truth.” He was a powerful fairy, who gathered the clouds on the peak. The fairy hares! on the hill were his pets, “no doubt,” and were often seen. He lived in the Sthrickeen,’ the big heap of stones on top. I heard from better educated (and therefore less reliable) sources that stories were told, the same as those embodied by Michael Hogan, “ the Bard of Thomond.” Hogan describes Donn, “in his chariot of meteors,’ defeating the Cratloe fairies, and carrying off the youne bride of Macconmara of Carrigogunnell; but the Ossianic flavour and bombast conceal the local mythology, if any. In a genuine (but probably late) folk verse, “Donn Firinne, Robert of the Carn, and Geroid Iarla, who vanished in the clouds,”® are the three chiefs of the Munster fairies from “ Carnthierna,” near Fermoy to the Shannon. Donn of Dotigh dwells in a large sandhill, in the golf links, near Lehineh, Co. Clare ; and I remember (in 1878) when people did not care to pass by night, and lights were said to be seen there. The last of the ancient bards of Thomond, Andrew Curtin (“fallen on evil days and evil tongues,” in the eighteenth century) prayed this Donn to take him as his servant. It was better than depending on the illiterate squires, though Curtin received much hospitality from the MacDonnells of Kilkee, who appreciated Ivish poetry ; but tradition does not tell that Donn granted the poet any favour. EpAoIn.—Though forgotten locally, the yellow-haired Edaoin, the tutelary goddess of Hogan Mogh Nuada, deserves mention.® She resided at Inis Grecraige or Beare Island, so their friendship may have been accidental. She rescued her protégé, and aided his escape to Spain by making the pillar-stones and rocks appear to his pursuers in the forms of his company. The deluded foe vainly broke their weapons on their supposed victims. ‘The story is evidently archaic, though we have it in a late and corrupt form. Elsewhere she appears with Cliodna and Aine as the “treasures of the Tuatha De Danann.” Fer F1.—I have said so much about the gods of Knockainey,® that I must 1 Hares were sacred in Gaul ; Boudicca brought one against the Romans. See Hibbert Lect. iv, p. 199. 2 ¢ Struadhraicin ” (FitzGerald’s ‘‘ Limerick,” ii, p. 382), ‘Sa specula or place for fire signals.’’ Lewis (Topog. Dict. p. 114) says it is on the site of an ancient temple. Illustrated by Dr. George Fogerty and described by Mr. P. J. Lynch, Munster Arch. Soe. i, p. 108. > Rev. Celt., iv, p. 191. 4 Mss. R. I. Acad., 23 M. 47, Folk Lore, xxi, p. 196. 5 **Magh Leana,” especially p. 31. 8 Supra, vol. xxxiv, pp. 50-56, 59, 60, 166 Proceedings of the Royal Trish Academy. confine myself to a short study of Fer Fi, son of Eogabal,* and a summary of the legends of his sisters, like that of the Nuadas. As we saw, Fer Fi thoroughly avenged the violence of Oilioll Aulom to Eogabal and _ his sister ‘Aine* In addition to luring his enemy’s family into civil war, he raided and burned Dun Claire fort, which deed keeps his remembrance green at Knockainey.s He had a holy mound at Sd Firat mhie Eogabail,' “to the south,” perhaps the most southern of the conjoined rings. He first discovered the Slige Cualann road, and was guarantor of the treaty of neutrality between Bodb Dearg and King Conn. Now Cacht, wife of Eogabal, was tutelary goddess of Fingin mac Luchta, an early king of Munster (cirea A.D. 130), who used to meet her at Kilfinnan, for Luachair Deadaid covered the district at Cliu, at least to the east of Cenn Febrat.6 Cacht broke the treaty by warning her favourite against every move of Conn.® Fer Fi’s mother is said in a late poem to have been daughter of Crimthann Nia Nair, King of Munster, “ B.c. 3 to A.D. 13,"7 who owned a fort of Cenn Febrat. Via Nair commemorates another tutelary goddess, ‘Nar, the witch of the Std mounds,”* who protected Crimthann on his naval expeditions from Howth;? but his fort there was not at the Baily (as O’Donovan decided), if it was in sight of Meath.” If Fer Fi (as seems probable) was Fer Fidail, son of Eogabal, he was instructed by Mananndn mac Lir, and used to bear a trident ; but, having abducted and accidentally drowned the Lady Tuan, the sea god slew him." Eter was his wife, and Emer was wife of ,Uainide."* Eogabal’s sons were Fer Fi or Fer J, Lu, and Fainle. The others, Fere, Fermait and Fer Fidail, are possibly alias names of Fer Fi. In the latest edition of the legends’Aine is also killed by Oilioll Aulom. A still later poem makes the latter straighten his bent spear-point with his teeth, one of which got poisoned with the '““Yew fork,” a suspicious name in view of the magic yew episode. Were the Knockainey gods connected with a sacred yew as well as with the hazel grove? * The name ‘Aine, Aina, or Ainia appears in Gaulish inscriptions, Holder, i, 71. *So Dr. Douglas Hyde informs me. I could not get any traditions on my visits to Knockainey. He has since published the legend in the Celtic Review in 1917. ‘ Onomast. Goedel., p. 599. ® On which lay their chief cemetery. ® Book of Fermoy (R.1.A. Irish Texts, Ser. 1). * Mesca Ulad, p. 53. Legends varied as to Crimthann’s death ; see Rev. Celt., xxiv, p- 176. * Coir Anm., p. 33, No. 106. ® «« Howth and its Owners,” Dr. F. E. Ball (R.S. Antt. Ir.), pp. 11, 12 ; Rev. Celt., ii, p. 86, and ‘‘ Eriu,”’ viii, p. 31. 1 Metr. Dind S., vii, Rath Esa, p.7; also Dublin Penny Journal, 1833-4, p. 60 ; the last is decisive for site of Dun Crimthann being at the mound at the martello tower in Howth. "™ Rennes Dind S., Rey. Celt., xvi, p 152. 2 “’Aine’s History,”’ Silva Gadelica, pp. 75-76, Wesrropp—The Earthworks, &c., of S. EB. Co. Limerick. 167 venomous blood of the banshee. Thus we have a complete evolution in the ‘Aine legends, in which she appears as goddess, princess, banshee, and poisonous monster ;' but to the peasantry only the first stage remains. The euhemerists were discounted, and ‘Aine remained beautiful, gracious, helpful, and deathless, as when the five Firbolg tribes adored her, before the coming of the Dal Cais. Tue KnockaInEy LicEnps.—There was no standard of pagan orthodoxy, and every tribe seems to have had its own recensions and divergent tales of the parentage and acts of the gods. Our study of Nuada, and W. M. Hennessy’s study of the war goddesses,? make this very plain. The tendency to make triads of the gods, even breaking up one into three gods, began early; the Gaulish carvings show the three birds of the war goddesses. The stories are kaleidoscopic, the same names and events reappearing in different combinations. In the ‘Aine legend* there were evidently two variants, one an inland version (perhaps of the Mairtene and Dergthene), making her and her relations children of Kogabal; the other a coast version (perhaps of the Corea Laegde), where the sisters were children or connexions of the sea gods.‘ (1) The Knockainey tales tell how Eogabal and his brother and family come from Uisnech ; the outrage of Oilioll Aulom and revenge of Fer Fi; the magic yew tree, and the semi-historic battles of Cenn Febrat and Magh Mucrima. In one ‘Aine is perhaps the wife of Dubthach, on Cenn Febrat, and her sister Aife, wife of Cain, on that hill. The earliest trace is circa A.D. 886, in Sanas Chormaic. “’Aine’s History ” has an early tinge; the others recognize the gods as wonder-workers, but lable to violence and death; the latest reduce them to fiends. (2) ‘Aine, Aife, Fer Fi, and Aillen of Sid Eogabail appear, so there can be no question of identity. The ladies are, however, daughters of Manannan, or his ollamh, or his father Ler, or his son Aillen, of Etar, or of Gailian; Etar is son (or grandson) of Ktgath. (a) One of the latter gods, usually Aillen or Fer Fidail, desires Mananndan’s wife. “Aine gives herself to Manannan, and obtains his 1“ Poisoned people” and weapons often appear, but perhaps metaphorically (Atlantis, iv, Coir. Anm., p. 307, and Wars of Gaedhil, p. 159). 2 Proc. R.I, Acad., x, p. 425 ; Hib. Lect. iv. p. 43; Rev. Archéol, xvii, p. 425; Prof, Anwy] (Celt. Rev., iii, p. 26) ; Dublin University Mag., Oct., 1834, p. 463; Rey. Celt., i, p. 39. 8 Supra, xxxiv, p. 55, and descriptions of Knockainey, p. 61, and Clogher, p. 63. 1 For this table (1) Silva Gad., vol. ii, p. 725; Metr. Dind., x, p. 229. (2) Battle of Ventry, Notes, p. 14. Duanaire Finn, p. 119. Introd. Feis tighe Chonain (Manannan and Aife) Duanaire Finn, p. 197; Gailian, p. 119, the ollamh ; Rey. Celt., xv, p. 331, Etar, Exchange of ‘Aine for wife of M., Silva Gad., p. 196 ; slaying of Fer Fi, Duanaire, p. 118; Rey. Celt, xvi, p. 152 ; Eriu iii, p. 151, Becuma. R.I.A. PROC., VOL. XXXIV, SECT. C. [25] : 168 Proceedings of the Royal Irish Academy. wife for her pining brother ; (0) or Etar’s wife is given in exchange; (c) ‘Aine’s sister Aife marries Lir (she is elsewhere different from ’Aine’s sister) ; (d) “Aine is daughter, not wife, of Gaidian, or Gailian; (d) the latter seduces ‘Aine’s foster-sister, Becuma, whom ‘Aine alone befriends ; (e) Aife and Fer Fi are pupils of Manannan, who slays the latter ; (7) Aife marries Lir. Outside the connected legends we have hints of many others. ‘Aine (daughter of Modann),' Lir, and Oengus of the Brug figure in one; Aife was daughter of Midir,* and the sisters vary as daughters of Eogabal, Eogamal, Durgabal, Lir, Gailian, Delbaoth® or Midir. Aife, Clidna, and Edaen, “ the treasures of the Tuatha Dé,’ are of Mananndn’s household, and get drowned when surf-riding at Glandore, whence Clidna’s Wave. Etar of Benn Etair (Howth), son of Etgath, pined away and died for love of ‘Aine.’ Aife, daughter of Delbaoth, of Lix’s household, is changed into a crane in the obscure “crane bag” story.° The Leabhar Gabhala gives Aidne (Clidna), Aife, and ‘Aine as daughters of Partholan; Lir’s children are probably the same group, Aodh, Aoife, and Ailbe, but they are daughters of Oilioll Aronn elsewhere, and the two first are successive wives of Lir.’ It is quite evident that three supernatural ladies, Aine, Aife, and Cliodna, were reverenced along the coast, at least from Glandore in Cork to Portacloy in Mayo, and that they give their names to the heroines of many divergent tales." No better example could be given of the impossibility of welding our material into a con- sistent whole, and no better excuse can be offered should one get lost in such a quagmire. A few closing deductions may be given. We have a mass of material from the seventh to the seventeenth century which can only yield results to very critical examination. The mere heaping up of extracts from every period can yield nothing but contradiction and confusion. Those who regard the least supernatural tales as the oldest, and reduce the stories to histories of mortals, are reversing the true process, as any attempt to arrange the sources by their periods shows at once the godlike god as the earlier. Those who shut their eyes to the results of Continental and British study, and believe the Gaulish gods to have been high kings in Ireland over a thousand years before Christ, ' Silva Gad., ii, p. 111. 2 Thid., p. 110; Tochmarch Etaine, Irische Texte, i, p. 127, Ir. Myth. Cycle, chap. xiv. 3 Ancestor of the Ernai and Muscraige ; perhaps this was their version. ‘ Silva Gad., p. 200. Another tale makes Clidhna daughter of Genand ; she drifts asleep in a bronze boat, and is drowned (Rev. Celt., xv, p. 437.) 5 Rennes DindS. Rey. Celt., xv, p. 380. ® Duanaire Finn, p. 118. * Magh Leana. * Leab. Gabhal., pp. 25-39; Atlantis, iv, p. 117. Wesrropp—The Larthworks, §e., of S. LE. Co. Limerick. 169 and accept the euhemerist “ Annals,” hardly deserve refutation, Such “irra- tional rationalizing ” as made the Tuatha Dé, flying on the wind to the north coast of Ireland, to be Scandinavians coming in ships, and perverted science by pointing out the forts and skulls' of what was a divine pantheon, is outside the true methods of study. Even a knowledge of Homer and Ovid might have shown our students that the heroes of Luachair and Muirthemne are replicas of the demigods who fought before breezy Ilion; that Cu Chulaind was a counterpart to Sarpedon and the divine Achilles, and that Lug and Manannan were but Apollo and Neptune in Celtic attire. In 4.D. 900, gods were recognized as gods; then the euhemerist movement by 1050 brought them down to dead kings and heroes and, later still, to magicians, protecting spirits, and family banshees, and at last to devils.° Only the peasantry were faithful to the spirit of the old tales; the god became a wizard and the druid a jester in literature. The gods “who o’er the Celtic roamed the utmost Isles’? were as nearly dead as the Aesir and Wanir. Yet the dead weight of the old literature kept back the expurgators. A “redeeming verse ” after a pagan legend, the praise of charity above keeping geasw, a verse on the Trinity after a poem to the “Seven Daughters,” reconciled the pious. It was as well, for there is less to offend a Christian spirit in the ancient tales than in St. Patrick’s abuse and brutal threatenings in the late popular Finn poems, so unworthy of the “humble and holy man of heart” of his own writings and of the early “ Lives,” and even stories. Ireland had no Saemund to give us uncorrupted tales of the old gods, but much survives before A.D. 1000 to enable us to “judge a people by their gods,” and get an all-important side-light on the brave and brilliant race, poets and missionaries and warriors, who evolved the gods in their own image before St. Patrick preached. One advantage the Irish mythology enjoyed—there had been no cruel struggle between it and the new faith.® The wise tolerance of 1 Sir W. Wilde, ‘‘ Boyne and Blackwater,” p. 239; Lady Wilde, *‘ Ancient Legends, Mystic Charms,” &e. (1887), pp. 353-7. * Magicians, Silva Gad., ii, p. 132, and often ; protecting spirits, Nar, ‘‘ Magh Leana” ; Cacht ‘* Book of Lismore” ; ‘Aine, supra, xxxiv, p. 59; Aibhin, Wars of Gaedhil, pp, 200, 201; Folk Lore, xxi, p. 26; cf, W. Stokes, ‘‘ Three Ivish Glossaries,”’ p. xxxiv. 3 Like Balor in Donegal, Ulster Journal Archeeol. (orig. ser.) i; maskers disguised as devils for Samhain night, see New Ir. Rey., xxvi, p. 145. King Cormac is slain by the siabra in old tales, ‘‘ by siabra and demons,’”’ says Keating (Hist. i, sect. xli), ‘*a devil attacked him,” say the Four Masters. The Book of Ballymote discusses whether the Tuatha Dé ‘ were diabolical demons,’’ or a human tribe. So late as 13817 the Cathreim Thoirdhealbaith makes the Badbh dwell in hell. So the Welsh gods became demons or fairies (Squire, Myth. Brit. Isles, ‘‘ Decline of the Gods”’). 4 Celtic Review, x, p. 263. ® Tn Scotland John Carsewell, Bishop of the Isles, denounced his flock for preferring tales of the Tuatha Dé Danann to the faithful Word of God and other cases. We have [26*] 170 Proceedings of the Royal Irish Academy. the Chureh (unlike its actions among the Scandinavians and Teutons)' passed no sentence of death on its rival. It had won “by the Word, not by the sword,” and, itself thoroughly Inish in a generation, could afford to let the S/d faith die out. Christians gathered at the old sacred places for consulta- tion* or pleasure; the pillar was marked with a cross or a holy name ;° the bile trees were left to flourish till cut by some hostile tribe, or felled by storm or by old age ;* the holy fire was tended by nuns; the wells rededicated and held in new honour; the old tales of the gods were told without offence before priest and pious chief. So, wisely and without hurry, the old faith was let die out, while, had other methods been adopted, the clue not only to Irish but to British and Gaulish paganism had been lost for ever. PART II. (F)—Tne Eartruworks. It is almost a relief to turn to tangible field-work once more. I may venture to tabulate the main types of the earthworks here. I have often opposed the confident statements of those who imagine they can determine the period and object of an earthwork by mere external inspection. Such theorizing (or rather allegation) is most dangerous to all sound archeology ; so we must try to avoid this great error. It is, perhaps, worse to take any fashionable theory in vogue outside Ireland, and apply it hastily to the antiquities of this country, playing the part of Procrustes in forcing facts to fit in it. In Ireland an error once in print in a journal can never be finally slain, for shallow beginners revive old theories and identifications, without (1656 to 1678) denunciations from Dingwall Presbytery against cattle sacrifices at an ancient temple, ‘* one god Mourie,"’ and ** Shony, a sea god,” the last for a good seaweed crop (Pennant’s ‘‘ Tour inthe Hebrides” ; Mitchell, ‘‘ The Past in the Present”; Squire, loc. cit., pp. 408, 412). “ The Celtic Church in relation to Paganism” (W. J. Watson, Celt. Rev., x, p. 263.) ' Elsewhere the Councils (of Arles, a.p. 452, Tours, 567, Nantes, 658, and Toledo, 681) fiercely forbade reverence of trees, stones, and wells, and punished those who prac- used it; cf. Olaf Tryggveson’s crusade against paganism in the Heimskringla and the various acts of Charlemagne. * A trench for a church was marked ‘in the name of the Lord of the Elements” at the green cf the Brugh for the first time in a.p. 499 (Rev. Celt. xxii, p. 415). * **Oulte des Menhirs,’’ De Jubainville (Rey. Celt., xxvii, p. 313, and xxviii, p. 222). For Cermand Cestach’s shrine at Clogher, Cal. Oengus, ed. Stokes, pp. 186, 187, 378. Plan, R. Soc. Antt. Ir., xxxiy, p. 320. Holy names on pillars (Tripart. Life, pp.90 n, 107). Some ogham stones have the god name after ‘‘ Maqi Mucoi” broken (Proc. R. I. Acad., xxvii, p. 333). For pillar marked with Christian emblems in France, see Life of St. Sampson. Cross-scribed pillars are numerous in Brittany as in Ireland. * Rennes Dind S. Rev. Celt., xv., pp. 419, 444; xvi, p. 277, yews, ash trees, and oaks. Westropp—The Earthworks, §c., of S. £. Co. Limerick. 171 testing the reasons for so doing, or rejecting where necessary, and seldom trouble to bring their reading up to a date much below 1840. I shall reserve the traditions relating to Knocklong and Dungrot, and confine myself to the earthworks near the first-named place, which complete the principal remains of the group from Kilfinnan to Kilmallock, Knockainey and Knocklong. I hope to deal later with the important remains of forts and mounds from Cromwell Hill to Dungrot. ‘The first was the site of a Std (though its god is, so far, undiscovered) in 1826. FitzGerald first noted “an inverted basin-like mound called Sighchann na Fionii” (Sided of the Fiana). The fine dolmen there has been already described, planned, and illustrated by Mr. P. J. Lynch and Dr. G. J. Fogerty. Forrs.—I have to use this unsatisfactory term for ring-walls and ring- ‘mounds, often residential, sometimes sepulchral, and never military; but the usage implies no dogmatic theory in Ireland. (1) Ring-walls of earth and stone, with (or without) a fosse and outer ring. (2) The D-shaped fort. (8) oughly square forts. (4) The crescent fort, abutting on a scarp of cliff, ike Dunganville. (5) Fortified spur,a mere variant of the last; e.g., one near Templenalaw. (6) The low platform fort, oval or round, with a fosse, and sometimes a ring. (7) The high mote, like Shanid Castle and IGlfinnan; the last with three rings and fosses, but no baily ; the former with a Norman keep and baily. Rarus.—The great ring-forts, like Dun Claire, Ballingaddy, and Ballinas- caula, are evidently residential. So, probably, are most of the platform forts, with the vast majority of the lesser raths. The platform forts of Magh Adhair, in Co. Clare, and Cush and Ballinvreena, in Co. Limerick, may have been originally sepulehral, or at least ceremonial. The first was, however, “besieged” in A.D. 877. The lesser house-rings may date down to (or after) the Norman settlement. One rath made about 1240 in Co, Clare. The “square” are probably late; but some in Europe date from the Bronze Age, and the making by the god Nuada of an earthwork with corners is mentioned in the “ Battle of Magh Leana.”” ConJOINED Rines,—These very interesting remains I hope to treat at some length hereafter. The occurrence of such at Knockainey, Clogher, Cooloughtragh (Temair Erann), and the ‘Oenach, near Monasteranenagh, mark them as for ceremony, and perhaps for sepulture. They have counterparts. Two, of different periods, conjoin at Tara,’ and others occur at Uisnech, and 1 Limerick, i, p. 406. * Loc. cit., pp. 1-4. $ The eastern, the so-called Teach Chormaic, I regard as the earlier, as the rings of the other join on to its perfect rings. [Since drawing this conclusion Professor Macalister’s monograph on Tara has been read. } 172 Proceedings of the Royal Irish Academy. near the great fair site of Cahermee, near Buttevant. So far I note two varieties, one of several disc barrows jor else mounds of the Rathnarrow type), as at Cooloughtra (five rings), Knockainy (four), and Doonakenna, near Barna Hill (three). The others are very low platforms. In Co. Clare there are two types—ring-walls, or rings of earth and stone, like at Teernea and at Liscullaun, near Corofin station ;' or a ring and crescent annexe (like at Clogher) as at Creevagh, Corbally, and Ayleacotty, near Quin, one at Moyfertagb church, and a marsh-fort at Lismehane,* Co. Clare, and Ballinbeg, near Aghada, Co. Cork. SerutcuraL Remaiys.—lIrish literature is rich in side-lights on such monuments. First, there is the duma, a tumulus, an outlook, or the raised site of a house. We have an example in the twin tumuli at Clogher, which work probably resembles the “Two Breasts of the Morrigu,” in the Great Cemetery of Brug*® When going over the site of Tara with Professor Macalister, we found what exactly tallies in site and character with the Treduma Nesi. O'Donovan was right as to its nature.‘ It was evidently three small mounds within an oval ring, partly terraced up; but the road has destroyed the middle mound. Such tumuli were probably (as in this case) connected with the gods. We read* how duma mounds were made over nobles, /eaca over heroes, fearta over warriors. We find elsewhere’ how a Jert of one door was for a man of science, one of two doors for a woman, and a mur, or ring-work, for those dying in a pestilence. When square or suitable stones were not to hand, square sods were used for the dartaire. The conjoined dise barrows at Cooloughtragh are evidently the joined tombs, traditionally, of Eithne and three other ladies, in the cemetery of the Ernai, “side by side.” Eithne was sister of MacNiad, and a daughter of Lugaid mac Daire, and was mother of Conaire by Mog Lama, and of Lugaid Laga by Mog Nuadat.’ Maer, who was buried near her, was a daughter of Ugaine’s son Fergus. The Knockainey rings were evidently connected with the gods ! Dr. George U. MacNamara has made mea plan which, with my own plan of Ballinbeg, I hope to lay before the Academy in a future study. * A groom named Clancy, long ago, was told by an old labourer not to exercise horses on it for fear of ill luck. * For two mounds at Da Chich na Morrigain (The Paps in Kerry) see ‘“‘ Burning of Finn’s House.” *O. S. Letters, Meath. Petrie fancied it to be three concentric rings. * First Battle of Magh Tured, mss. T.C.D., 21.2.17, p. 91. ® Book of Lecan, f. 258; Keating’s ‘‘ Three Bitter Shafts of Death,” for ‘‘ small raths of the claidhe”’ used for burial ; see ‘‘ Tract on the Cemeteries” ; Book of Lismore ; (Rev. Celt., xxvii, p. 328) ; Story of Etain, L. na hUidhre, names the fert of Dughis ; see also many notes in Agallamh and Dind Senchas. In Tochmarch Feirbe fifty persons are buried in one duma; in ‘‘Cath Crinna” (mss. R. L. Acad., 23k 37), two druids are buried in Dumha na ndruadh. See also Cath Finntragha, p. 87. 7 Silva Gadelica, ii, p. 524. Wesrropep—The Earthworks, &c., of S. EB. Co. Limerick. 173 —Uainide, with the northern; Fer Fi, with the southern; and, perhaps, Eogabal and ‘Aine with the rest." The brothers of Medb were buried in a mur at Ratheroghan, and, in the sixth century, St. Senan was buried in a dere or fert, a square enclosure with upright stones, such as was connected with his name on Iniscaerach (i.e. Mutton Island), Co. Clare. The shield of Cu Chulaind? was a dise barrow, level with the field, and with a small mound in the centre. Bowl barrows occur at Cush (5) and Clogher; dise barrows at Cush, Clogher, and Ballinastona, in Co. Limerick; Lislard, Lishaun, and Tyredagh, and George’s Head, Kilkee, Co. Clare. (4) Rownded low mounds, like Rathnarrow. One remains at Lisdoonvarna, Co, Clare; others at Kmockainey and Ballygubba, Co. Limerick. (5) Cairns in a ring—Knock- ainey, Knockadoon. I have not found a kerbed cairn in Co. Limerick, such as we find at Leana, Poulawack, and Slievenaglasha, Co. Clare. There are traces of large cairns at Knockfierna, Carnarry, and at Seefin. (6) The Limerick dolmens are chiefly cists, the Duntrileague one is complex. Kyock1ione (Ordnance Survey No. 40). A fine group of earthworks extends from Clogher through Knocklong, and down the valley of the Saimer, or Morningstar. There are no traces of earthworks on the ridge of Knocklong, whose fifteenth-century castle of the Hurley’s and the graveyard are so conspicuous from the railway from Dublin to Cork. It is the ancient Drom Dambhgaire, the scene of the very mythical defeat, or rather flight, of King Cormac,’ about A.D. 230. I must study its legend hereafter, and only give a brief sketch of its history and traditions, The O’Hurleys built the castle late in the fifteenth century, after 1460. The records only begin in the reign of Elizabeth in the “Fiants” of 1568, 1570, and 1584, as Loinge and Knockneloinge. The Hurleys held it till Sir Maurice Hurley lost all in the civil war of 1641-51. His confiscated manor had “a ruyned castle, a mill, two fairs, and Courts Leet and Baron.” His lands extended from Dunmoone and Mitchellstown down to Startin (Searteen) and Ballinalanga on the borders of Co, Tipperary. In 1853 it was said to take its name from Jong, a ship in which Hurley used to sail from it to Emly. Itis true that there was a lake at Emly in A.D. 896, named in Cormac’s Glossary, and there are traces of several other lake-beds; but no 1 Supra, xxxiv, p. 62. 2 Metr. Dind S., viii, p. 17. Similar to this may have been Sciath Nechtain (Wars of Gaedhil, p. 21, A.p. 847), at Skea, Co. Kildare, and Sciath gabhra, or Skea, Co. Fermanagh. 3 Forbais Droma Damhgaire, Rev. Celt., xv, p. 441 ; Keating's History (Ir. Texts Soc.), ii, pp. 319, 320; O’Curry, ‘‘ Manuscript Materials,” p. 271. 174 Proceedings of the Royal Irish Academy. continuous sheet of water could have reached from it to Emly in historic times. Windele found the old legend still remembered, how a “king of the Dandonians” and his army suffered from thirst, and his druid “shot a sleagh,” and where it pierced the ground the well of Curraheen, on the north-west slope of the ridge (Slievereagh), broke out.' Mr. Molony, of Hill House (to whose kindness and hospitality I am indebted), tells me that, when digging a fence above the quarry on the west slope, a layer of bones of men and horses was found. Here John Windele, in 1853, found “a limestone pillar, or dal/dn, 4 feet high and 13 inches thick, on the edge ofa precipice.”* Mr. Molony also tells me that long ago a cave was opened between the castle and the graveyard, and silver candlesticks were said to have been found, and long preserved by a Mr. Ryan. The railway cuts the group of forts in two. Unfortunately, Windele’s untidy method of making notes on any loose bit of paper, and hardly ever writing out a full account while his memory was fresh, deprives his jottings of much of their value, and the Ordnance Survey Letters, as usual, ignore them altogether. I take them from the north, southward :—(1) The northern mote is 8 feet to 11 feet high, circular, flat, or slightly hollowed, on top, and 51 feet across. The sides are steep, especially towards the north, with hawthorns on the sheltered side to the south-east. The fosse is 12 feet wide and 2 feet to 4 feet deep, wet, and full of “ flaggers” (yellow iris), with a trace of an outer ring, 10 feet wide and 2 feet high, to the south-east. This, I presume, is Windele’s “ Liss of Knocklong West,” described as a mote “20 feet high,” with a deep fosse, 20 feet wide and 56 feet across the top, which is slightly hollowed. It lies beside a little rivulet, in wet fields, now drained. (2) In the same marsh, southward, is a curious platform fort and a ring-fort in Knocklong Townland. The first consists of two rectangular platforms, in line, north and south, 5 feet to 6 feet high and 66 feet wide, within a fosse 12 feet wide, now nearly filled; the two divisions had banks round the tops, and have a fosse 12 feet wide between. The northern is 70 feet long, much levelled to the north-west: the southern is 54 feet long. (3) At 63 feet to the south-west is a marsh fort, a ring, 66 feet inside; the rampart is 6 feet thick and high, the fosse 12 feet wide and only 2 feet to 3 feet deep, with an outer ring 3 feet high and 6 feet thick. It was nearly concealed in tall meadowsweet, loose- 1 Fiants Nos. 1765, 2472 ; Desmond Roll, 1584 ; Civil Survey, vol. xxv, p.9; Windele’s ‘*Supplement,” i, pp. 526-530 ; Sanas Chormaic, p. 93. 2Mss. R. L Acad., Windele’s ‘‘Supplement, i, pp. 526-530. Did the pillar commemorate a battle ? cf. Da Derga’s Hostel (Rev. Celt., xvii, p. 169), ‘‘a pillar stone for a rout, a cairn for a destruction.” Wesrropp—The Earthworks, &c., of S. B. Co. Limerick. 175 strife, and flaggers on my visit. Windele describes it as “a mote 13 feet high, with a cave, in which a dog was lost.” This opening is not to be seen, and though he places the mote near the oblong platform, he probably means the next mound. (4) The mote near the Elton road is 12 feet to 14 feet high ; it has a dry fosse 15 feet wide, and a sort of ledge, such as one finds in bell- barrows, to the east, round the foot of the mound. The flat summit is from 64 feet to 66 feet across, the base about 90 feet. Many hawthorns grow on its side. There is no outer ring. A long circuit through the village brings us past (5), a low liss, a thicket of thorn-bushes, with a fosse and inner and outer rings, to the north of the railway, east from the station. We eventually reach a group with two good forts south from the line, from which they are well seen, with their noble background of mountains. Owing to the increased tillage (from the German submarine campaign), many fields round them were broken up; I carefully, but vainly, looked for sites of hearths or objects of antiquarian interest, but saw none. (6) The fort nearest the railway is a perfect little oval mound, 11 feet to 12 feet high, 42 feet across north and south, by 30 feet east and west on the top. The fosse is 15 feet wide, 4 feet to 5 feet deep, and partly wet, with no outer ring. (7) Another fort, which I failed to find, and could not see from the railway. (8) The most southern, and finest, mote, near the road from Knocklong ridge to Aghadoon, is on the summit of the plateau; it has a beautiful outlook to the Galtees and Slievereagh. It isa very perfect mote, 15 feet to 15 feet high over the fosse, 40 feet to 45 feet across the top, and 70 feet at the base, with steep sides, and thick hawthorns growing on it to the south-east. The fosse is 16 feet wide in the bottom, and over 6 feet deep, being still wet. The fact that so many raised flat-topped forts remain where no castle is recorded is noteworthy, there being, as we see, four at Knocklong, and nine near it at Atheneasy, Aghadoon, Ballinvreena, Cush, Raheenawadra, Ballinscaula, Bulgadin, Glenbrohaun, and Rathtany.2 None of these occur in the sanctuary cemeteries of Oenach Clochair and Knock- ainey, though the third and fourth are near Cush. This seems to imply that they are not necessarily either burial tumuli or feudal castles, for what need of four in one townland? and the number precludes the idea of their being inauguration mounds, like their congener at Magh Adhair. I can only draw the conclusion (so strenuously denied) that such flat-topped forts (and the remark applies to Pallas, and to those in Co. Clare, Lisnagry, Lugadoon, Killilagh, Moyarta, Lisnaleagaun, and Kiltinnaun) are Irish residential “forts.” The bearing of this on Shanid rath is also very clear. 1 Plans of Nos. 2 and 3, and sections of 4 and 8, are given, Plate VI. 2 Ratheany in Inq. post mort. of Tho. de Clare, 1287 ; Rathtany in Memoranda Roll Exchequer 1317, m 82, as to succession of his son Richard. R.I,A. PROG., VOL. XXXIV, SLOT. C. [26] 176 Proceedings of the Royal Irish Academy. AGHADOON, Knocxrorry (0. 8. 40).—The Saimer, to judge from its old channel, was once a goodly stream, and if it be the Saimer of a list of the chief rivers of Ireland, got more than its share of recognition. Now, the great rain-collecting forests being long felled, it, like the Cammoge, has dwindled to a brook. It divided the Ui Fidgeinte from the Arada before the Dal Cais intruded. The ford Aghadoon (Athnaduin), also called “ Doon Ford,” is marked by the old road from Knocklong to Kilmallock on either bank. It is evidently the Athdunbari,’ named, along with Laythyralaw (or Templenalaw, on Slievereagh) and Garthagriffin (Ballingarry in Coshlea), and was named from the Barrys. It is most probably the ford at the earth- work of Raithin an Imaraigh, “the little fort of the contest,’ where, in the “Battle of Knocklong,” Colga, King Cormac’s chief druid, was slain by the superior magic of King Fiacha’s druids, who turned his “handstone” into an eel or serpent,? for the ford lay west from Knocklong and had a little rath near it. The utterly mythical nature of some Irish Sagas never affects their accuracy in topography. The Doon is a low mote in Knocktorin = a shapely mound, carpeted with ferns, violets, and primroses, and planted with sycamore and beech. Through these trees it has beautiful views of Slievereagh to the south, and a glimpse to the north of Thountinna, over Loch Derg. The mound is well seen from the train, between Knocklong and Kilmallock, to the south. It is 16 feet to 18 feet high, the top girt by a low ring, 8 feet to 10 feet thick, and rarely 3 feet high. The platform inside this measures 60 feet across, or 123 feet at the base. There are traces of a fosse, but too defaced to be measured.! The old road runs from it to the ford, westward down the slope. ATHENEASY (O.S. 40).—Like Aghadoon, the next ford down the stream to the north of the railway has got a high fort near it, but on the other bank. The name Ath na nDeisi refers to a tribe, the Déis, denizens of the Tara district of Bregia in Meath, who are said to have fled for refuge to Oilioll Aulom in the late second century. Under his gis, one section settled in the present baronies of Decies in Co, Waterford, the other in Deisbeg, or Small County, in Co, Limerick. It is called Athenysy in a papal letter of 1260. In 1335, Elizabeth Milton, widow of Walter de Bermingham, had owned Athnedess. The church of Aghnedesse was robbed by John Staloun ; ' Plea Roll, No. 22 of xxv, Edw. I (1296), Pars iii, m 48, and No. 42, anno xvii. ? Similar tales of great eels are found near Loch Gur, Rey. Celt., vol. iv, pp. 171, 186. In the Tain bo Cualnge (ed. Dunn), p. 161, the Morrigu turns into an eel to wind round Cu Chulaind at the ford ; cf. also an eel peist in Co. Kerry (mss. R. I. Acad., 12c 3 (13), p- 147 ; Roy. Soc. Antt. Ir., xxv, p. 74), and Feis tighe chonain, Introd., p. iii. 3 Cnocktoren, mearing with Dunmoone and Elton, Small County (Civil Survey, p. 10). § Plate VI. Wesrropp—The Earthworks, §c., of S. E. Co. Limerick. 177 Athnedisse rectory is again named in 1393, and the manor of Andesshe was held by the Rolley, or Raleigh family in 1408 and 1424. It is called, in 1410, Athnedisse, or Beallathenesigh (the latter being the ford), which retained its old name, Beul Atha na nDeisi, in 1579.1 The thicket of trees, and bushes in the mound, keep the fort unknown to many who frequently drive past it. It, like the Doon, lies not far from the village of Elton. This name is supposed to be very modern, but it is found as “ Elltown, the pro- perty of Sir Edward Fitton,” in 1586; two years later, James Fox of Elton claimed the land which Fitton held as patentee. Moriertagh O’Grady (I presume as Fitton’s tenant) held Elltown in 1610, and it is mentioned very frequently in the Civil Survey in 1655. The mote is from 18 feet to 20 feet high, surrounded by a fosse 18 feet wide, and rarely over a yard deep, with a stream running into it to the south-west. The mound is 51 feet to 54 feet across the platform, and 104 feet at the base; it - is well preserved and the sides are steep, but it is overgrown with hawthorns to the east. BALLINASCAULA (O. 8. 40).—Of a different and more interesting type is the mote of Ballinascaula. On a lesser scale of height and massiveness, it still resembles Kilfinnan in having a raised mound girt by three rings. It hes in a marsh, not far to the north of the railway, and between Bulgadin and Atheneasy. It is locally supposed to mean “fort of the heroes,” or “fort of the clouds,” being on the map “ Ballinscaula,” but locally “ Ballinascaula,”* Like its neighbours, none of its early records remain. In 15835, Gerald mac Thomas, alias “Tonboy Reagh,” a Geraldine, held Glenlarhy, Ballin- wryny (Glenlara and Ballinvreena on Slievereagh), and Ballinskaly. It was granted, May 14th, 1588, to Richard and Alexander Fitton as part of their demesne, “ Phitton’s fortune,” long forgotten, along with it, the above lands, Coch (Cush) and Ballenvistellane down, or Mitchellstowndown; the grantors were pledged to erect houses for twenty-three [nglish families; Mahone mac Teige held Ballynscholly under Sir Edward Fitton. The Civil Survey, 1655, gives “Ballinscala, half a townland, mearing on the north with Gormanstown, and with Bulligidyn-Eady to the west and north-east.”* No tradition seems to attach to the forts; Mr. Thomas Bennett, of Summerville House (the owner), tells me that there were traces of other small forts round it, but they were levelled at various times in farming the place. 1Cal. Papal Letters, i, p. 370, vol. iv, p. 458, Close Roll, viii Edw. III, 146, Plea Roll, No. 123 (1318-20), Pat. Roll., x Hen. V, pars. 2, No. 24, Ann. Four Masters, 1479. 2 Plate VI. 3 Ing. R.L.A., i, p. 75. Ing. Exchr., Nos. 11 & 12. Proc. R. I. Acad , xxvi, p. 189. Fiant 5179 & 5032. Civil Survey, xxv, p. Ll. ropes Tombstones inscribed with craft emblems are comparatively rare in Munster, though they are of frequent enough occurrence in other parts of Ireland. The writer remembers no other example in the present barony, but in the adjoining Imokilly barony there are at least two examples—if the horseshoe on the supposed Smith monument in Cloyne Cathedral be a craft emblem. The other example is in Aghada old graveyard, and will be referred to again and described under Imokilly Barony. Power— Place-Names and Antiquities of S.E. Cork. 209 L.H.S. MARIA. ANO. DNI. 1614, RICARDUS [B] A DEMO... . DON. The initial letter of the surname is doubtful. S.DD. Poll Buidhe—* Yellow Hole”; a field, close to the chapel, so called from a pit which yielded a species of marl or yellow earth. “The Warrens.” An Carraigin—*'The Little Rock”; a field. “The Pedlar’s Rock”; this is a cliff overhanging the river; from it a wandering chapman fell into the water, and was drowned. Poll a Tairbh—*‘'The Bull’s Pool”; a deep hole in the river; perhaps a bull was drowned here. Poillin Bhriain—“ Brian’s Little Pool”; another river hole. Paire na Cloiche—“ Field of the Stone”; the stone is the dallan alluded to above. Paire a Ghaid—“ Field of the Withe”; possibly from an execution by hanging; for gad is sometimes used to signify a halter. The present field has an uncanny reputation. Jack-o-the-Lantern, or some allied sprite, made the place a scene of his nocturnal pranks; see under Gortagousta, par. Carrigtohill, antea. Paire a Cliamhain-Isteach—“ The Son-in-Law’s Field.” Cliamhain-Isteach is a son-in-law who comes to live in his wife’s house; cliamhain is a son-in- law simply. BALLYHAMSHERRY, Baile a Hampshire—‘ Hampshire’s Homestead.’’ The family name, Hampshire, is not now known locally; but in 1366 one Thomas Hanser! was a tenant of church lands in this locality. Area, 383 A. Ballyhamshier (D.S.R.). On the townland, in a field by the roadside (Moore’s farm), is a pillar- stone, of Castlelyons type, measuring 6 feet by 5 feet by 2 feet; the pillar is of limestone, and on its south face are some natural markings, suggesting the imprint of four gigantic finger-tips, and indeed believed locally to be such— proof conclusive that the stone was once a giant’s plaything! S.DD. Pdire a Dall4in—< Field of the Pillar-stone”’; this is the field on which stands the monument just described. 1 Rotulus Pipae Clonensis (Caulfield), p. 30. Probably Hanser in the Pipe Roll isa mistake for Hamsir (Hampshire). [80*) 210 Proceedings of the Royal Irish Academy. Crosaire na Caillighe—“The Hag’s Cross-Roads.” The cailleach in the present instance was no female of supernatural origin or powers, but a poor, unfortunate old creature, whose dead body was found here by the roadside upwards of a century since.’ Bealach Abhann—“ River Passage’’; a ford in the Bride river. The place is now spanned by an iron footbridge; but wheeled vehicles still use the ford, as did their predecessors in the spacious days of Carthage and Colman. Paire a Leasa—“ Lios Field.” ‘There is no lios on the townland at present, but the name shows that things were not always thus. BA.iyoran, Baile Uf Odhrdiin—“ O’Horan’s Homestead.” Area, 575 A. Ballyorane (Deps. 1652). Towards north side of the townland, on James O’ Brien’s farm, is a ceall, or presumed early church site. ‘There are no remains of the church, nor even traces or tradition beyond the field-name—An Ceall. S.DD. Bothairin an ‘Atha—“ Little Road of (to) the Ford.” “The Long Quickeen,” a field. Iam unable to explain the name; possibly it has reference to a former quickset fence, or it may be coicfn, a “ cock’s comb.” “The Cautheach ( Caitéach)’”—* The winnowing place.’ Sean Abha—‘ Old River”; the brook or stream which drains the town- land into the Bride. Corrach na Druimfhinne—“Swamp of the (Legendary) White-backed Cow.” BALLyROBERT, Baile Roibedrdaigh—Idem. According to a note in the O.S. Field Book the castle and townland derive their name from a Robert de Barry, by whom the stronghold was originally erected in 1325 as a frontier defence of his barony. Area, 169 a. Ballyroberts, ats Robertstown, ats Ballygobnett. (Deed of sale, 1702. Vid. Cork Archaeological Journal, vol. xxii, p. 102.) Portion of the townland runs into the neighbouring parish of Knock- mourne. On the townland are a ruined castle, a Holy Well, and a lios of unusual character — all on Mr. Mackey’s farm. ‘The castle stands in the farmyard; its remains consist of a keep, or rather peel-tower, minus its south side, which has been blown completely out. In the second story are ! Ordnance Survey Field Book. * Winnowing was done in the open, and was generally the work of women. The venue was a slightly elevated spot unsheltered by trees or fences, and the time—a day windy but fine. Light sieves were the only implements or machinery used. The operator, facing the wind, raised the sieve of corn to height of her shoulders, and allowed the contents to fall in uniform stream to the earth. Needless to add, the ground was covered with a winnowing sheet. Powrr—Place-Names and Antiquities of S.E. Cork. 211 traces of vaulting. A curious feature is the outward splay of the opes; evidently the design was defence rather than light. Ballyroberts was a Barrymore castle in the seventeenth century, and as such it was besieged in 1645 by the Earl of Castlehaven.’ Probably destruction and final abandon- ment of the castle date from the Confederate period. Thenceforward, owing to the development of ordnance, there would be but little inducement to castle-rebuilding. There is a sketch of the castle in the Windele mss., 12 J,11. R.I.A. p. 239. Close to, and partly underneath, the east boundary fence of the townland is a rather noted Holy Well, popularly dedicated to the Blessed Virgin under the title of Lady’s Well, with devotions on August 15th. The Pattern had degenerated into a scene of drunkenness and faction- fighting, till it was finally abolished, nearly a century since, by Mr. Mackey and the local clergy. At date of my visit the well was in a very neglected condition—dirty, bare, and nearly empty. Close to the well, separated, however, from the latter by a fence, and, therefore, within the townland of Grange, is a pillar-stone, at which “rounds” are also made! perhaps it would be more correct to say that the well “rounds” were prolonged to the pillar-stone.? The lios is a remarkable specimen, covering about an acre of ground, and defended by no fewer than three perfect concentric ramparts, of no great height or strength. This fine monument will be found within Mr. Mackey’s lawn close to the entrance lodge. Near the lios there was found in 1838 a fibula of pure gold, which was sold in Dublin for £27. S.DD. Poll Buidhe—“ Yellow Hole,” from the yellowish, marly soil. The name is applied to a field within which is a fox-covert. Tobar na Faille—“ The Cliff Well,” by the riverside. Clais na gCoinini—* Trench of the Rabbits.” Paire na Mona—“ Turf Field.” BALLYTRASNA, Baile ‘Trasna—“'lhe Homestead Across (the Stream).” Area, 419 A. Ballytrasney (Inq. Car. I.). There is a single circular lios of medium size on Ivis’s farm. S.DD. Collatrom (O.M.). This is the name of a stream which forms the eastern boundary of townland, parish, and barony. ‘There is a place of the same name in the parish of Kilbrogan, near Bandon, and in the latter case Canon Lyons® renders it Caladh-Trom, ie., “Holm of the Elder- bushes.” Prof. O’Donoghue thinks it most probable the name is Caladh. 1 Vid. Waterford Archaeol. Journal, vol. x, p. 9. * See under Grange, infra. 3 Cork Archaeological Journal, vol. ii (1893), p. 9. 212 Proceedings of the Royal Irish Academy. trom, while Mr. P. M‘Sweeney suggests Caladh-tirm. On either supposition the stream would derive its name from the river-meadow. i Tobar-Nat—“ Nat’s Well.” An Seairt— The Thicket,” a field. Barr a Bhaile—* Village Summit,” another field. Corin, Carn—“ Sepulchral Stone-Pile.” Area, 714 4. The name-giving carn, now unfortunately ruined, is itself Carn- Tighearnaigh.!. Tighearnach may have been the great man whom the monu- ment was intended to commemorate. Alas, nothing of him, beyond his bare name, survives. Perhaps the assertion is too sweeping; within the tumulus were found, in 1833, two fine burial urns of Bronze-Age character. One of these was broken, and the other passed into the hands of the Rev. Joshua Brown Ryder. ‘lhe ultimate fate of the surviving vessel I do not know; but, fortunately, a good drawing, or rather an engraving, of it survives. It was 54 inches high, by 3 inches in diameter at base and 5} inches at mouth, and was furnished with a conical cover. ‘The cairn, before its destruction, is described as somewhat irregular in outline, 19 paces in circumference at base, 26 paces in ascent, and 11 paces in circumference at the top, where it was crowned by a stone pillar, 8 feet high. Croker adds that, surrounding the cairn at a short distance, there was a circle of cyclopean stones. Local folk-lore connects our carn with the prince (in this case, Robeard-a’- Chairn) for whom death by drowning was foretold. The anxious father, to remove all danger, proximate and remote, of the prophesied end, had the child removed for nursing to the summit of this wild and waterless peak, where, at the height of 727 feet, a residence was built for him, Lut futile all efforts to escape the stern edict of fate; the child met the end decreed in a basin of water! Croker has embodied the story, with yet another tale of Carn-Tighearna, in his “ Fairy Legends of the South of Ireland.” On Corrin townland are likewise two, more or less holy, wells; also the sites of two lioses—one each on Coghlan’s and May’s farms. S.DD. Tobereendowny, (O.M.), ‘Tobar Riogh an Domhnaigh—“ The Lord’s (Jit. King of Sunday’s) Well,” i.e., well at which devotions were performed on Sundays. ‘This is on Coghlan’s farm, and “rounds” are still occasionally made. Tober na Naomh— The Saints’ Well”; notwithstanding its name, the sacred character of this well is somewhat doubtful; at any rate, the well is" not so emphatically “holy” as its sister spring, just described. 1 Carn Tigernaich in territory of Fear Muighe Feine (O’Clery’s ms. Life of St. Finnbarr). 2 Windele mss., 12, I. 11, R.I.A. Powrr—Place-Names and Antiquities of S.L. Cork. 213 Loch a Phréacain—“ The Crow’s Pond.” This is on the southern boundary of the townland. Bothairin an ‘Atha —* Little Road of the Ford”; it crosses Corrach-na- Druimfhinne, and leads, incidentally, by and to the ceall on Ballyoran. “The Priest’s Road”; it runs up the south-east side, or shoulder, of Corvin Hill, The name is probably due, as in similar cases, to the fact that construction of the road was owing to efforts, or representations, of the parish priest. Corrach-na-Druimfhimme (see under Ballyoran, supra), This is an exten- sive bog, lying on east side of the townland and adjoining Ballyoran. Local Irish speakers had a legend of the name-giving Cow, but, unfortu- nately, I failed to find an Irish speaker who remembered it. Legends retailed in English for edification or delectation of the Sasanach are mostly spurious. Associated with the swamp was likewise a second legend of a supernatural, or magic, eel.’ It is believed, too, that the bog was anciently a lake, and that the latter was drained to defeat the fate predicted for the prince of Corrin legend. In this bog, Mr. J. W. Sherlock of Fermoy discovered, about 1843, the entire skeleton of an elk, which he presented to Lord Mountcashel. In the same place, or neighbourhood, a Rey. Mr. Mockler found a second specimen?. The Leaca (Leaca—a glen-side) ; a field. Dererpark. Pail an Fhiaidh— ence of the Deer.” The form is unusual ; of. Deerpark, parish of Lismore (Co. Waterford); Deerpark, parish of Carrick- on-Suir (Co. Tipperary), &e.5 Area, 336A. On the townland were three lioses. Two—of circular outline—have been utterly destroyed, but a small oval specimen survives. On the townland is likewise the site of Killawillin early church and graveyard—tfrom which it is pretty evident that Deerpark is a detached portion of an ancient, and larger, Killawillin. The present townland was the deer enclosure of the Earls Barrymore, whose principal residence was at Castlelyons. ‘lhe Ceall site is marked by a large ash-tree on Patrick Shinnick’s farm. S.DD. Carraig na Cille. “The (Early) Church Rock”; site of the Ceall alluded to; there are no remains. 1 Since writing the above I have found the legend in the Windele mss., RI.A., vol. 12, I, 11, p, 243. The Carn was long ago called Leaba na Lun, from ‘The Lun,” a wonderful animal which abode there. This monstrous creature had four legs, great prominent eyes, and a tail, a single lash of which would uproot an oak. The monster turned itself into a piast for the purpose of milking the Drimin at night. Finn and his hound were, however, sent for; they came, saw, and conquered ; valiant Bran attacked and killed the Lun. 2 Windele’s Notes, R.I.A., 12, I. 7, p. 572. ’ Place-Names of Decies, pp. 27, 246, 260, and 266. 214 Proceedings of the Royal Irish Academy. “The Weir Hole”; in the Bride river. GLENAROUSK. Gleann na Ruse. Meaning somewhat doubtful. The dictionaries have not the word Résc, which is also unknown in the colloquial Trish of the locality. O’ Donovan, however,' renders the word—a marsh or fen. Canon Peter O'Leary, in whose parish the townland is, thinks Ruse = Ruadh Uisgue (i.e. Red Water). In older Irish rise = a push, drive, rush (of water). ‘lhe word ruse in the name Drumrusk, Co. Waterford, I have?— following O’ Donovan*—rendered “ fleeces.” Ruse occurs again in the name Roosea, parish of Tubrid, Co. Tipperary ; and in Roosca Cross, near Cloyne. Area, 1059 a. On Hayes’s farm, towards west side of the present extensive townland, and in a field called Péire-a-Leasa, is a small pillar-stone, known as “The Dallan.” In a neighbouring field, on same farm, is a well to which a repu- tation for sanctity in some degree attaches. On this farm there are likewise three sinall circular lioses or traces of them. S.DD. Pope’s Well (O.M.); near south-west angle of townland. Jarley Bog (O.M.); near north-east angle. A cannon-ball, weighing 7 lbs., was found here about seventy years ago. Lady’s Well (O.M.). No “rounds” are made there now. : Jawnanimirish (O.M.), Ban an Imris—“ Field of the Battle,” from a skirmish here between the Cromwellian and Irish forces in which the latter were defeated. “The Barrack Field”; so called from a cluster of labourers’ cottages. Paire a Lingthe—“ Field of the Leaping”; from athletic contests of which it was once the scene. Pdire na Stagin. I do not know the force of the qualifying term in this or similar names. (See under Ballynagarbrach, parish Caherlag, supra.) Paire a’ Leasa Uf ‘Og4in—* O'Hogan’s Lios-Field.” Paircin a’ tSil Fhéir—* Little Field of the Hayseed” ; or, perhaps, P. a tSoiléir—* Field of the Cellar.” PAire na Cloiche—* Field of the (Pillar?) Stone.” As the stone has been, loug since, removed and destroyed, we can only conjecture its character. “The Poteen Well.” An enterprising, but not law-loving, band found the water here suited to their illicit purpose. FarraN, Fearann—‘ A Farm.” In place-name usage I find the term fearann often or generally applied to Church-land—a glebe on the tenanted land of an abbey. On O'Brien's farm is a ceall or early church site, close to Bothairin an “Atha. Area, in two divisions, 207 A. 1 Ordnance Survey Field Books. * Place-Names of Decies, p. 212. ® Field Books as above, Co. Waterford, Powrr— Place-Names and Antiquities of S.E. Cork. 215 S.DD. Shanowenadrimma Stream (O.M.), Sean Aba na Druimfhinne— “The White-backed Cow’s Old River.’’ The Ordnance Surveyors apply this tautological name to the stream which forms the eastern boundary of the fownland. “The Lios Field,” on Broderick’s farm, where there is no lios now. GRANGE, Grdinseach—Idem. A grange is more especially the out-farm of a Religious House. Area, 537 A. On this townland—just within its boundary fence, near the south-west angle—is a pillar-stone of the same type as the Ballyhamsherry monument. The present dallan is deeply embedded in the earth and is, in some way, associated in popular estimation with the Holy Well on Ballyrobert. A few years since there were three lioses, but only one—a large specimen—now survives. On Andrew Meade’s farm in low-lying, wettish land, are two or three folacht-fiaidhs or prehistoric cooking places. 8.DD. “The Castle Field.’ ‘This is on Andrew Meade’s farm. Cannon- balls have been found here; also the foundations and other remains of ancient buildings. Paire na Bé Mairbhe. ‘The Dead Cow’s Field.”’ Kitcor, Coill na Coradh (or, na Corra)—* Wood of (beside) the Irrigation Dam (or-‘of the Round Hill’).” According to Canon Lyons, “ Corradh”’ signifies a homestead. Area, in two divisions (including a considerable portion of mountain), 1427 a. Jalleurr (D.S.R.); Kilnacor, Kilnecur, and Coilnecurra (ancient authorities quoted by Windele—wms. R.LA. 12. I. 11, pp. 191, &e.). This townland lies on, or contiguous to, the southern boundary of the parish, and on it are two cealls, together with the site of an ancient castle. The beauties, glories, and memories of Kilcor were celebrated in a popular song, which began :— “* Coill na Corra cois Brighde.” $.DD. Tilcor Castle (O.M.). According to local tradition, this was a castle of the O’Briens, who held under the Barrymores, but were obliged to return to Clare (by transplantation) in the middle of the seventeenth century. The tradition almost certainly embodies historic fact. Compare the document quoted under Pellick, below. Cill na Corra. The ceall site, on Pyne’s holding, close to the castle. Paire a tSéipéil—“ The Chapel Field,” on Hegarty’s farm. The name suggests merely a chapel site of the Penal days, but the tradition of a grave- yard attached points to an early Celtic church site. 1 Cork Archaeological Journal, vol. ii, pp. 169, 170. R.I,A. PROC., VOL, XXXIV, SECT, 0, [31) 216 Proceedings of the Royal Irish Academy. Méin Ruadh—“ Red Bog,” a subdivision, containing 53 acres. Poll a Tairbh. See under Ballyarra. “Putty Hole,” a hole in the Bride River: probably from pota, a pot. Cnoc a Mhadra—* Hill of the Dog (wolf).” Ocrean—probably, Eo Crion (i.e. “ Withered Yew-tree ”), a field. Piire na gCloch—* Field of the Stones.” The stones here are not dallans but surface boulders and smaller stones, in which the place abounded. Paire a Tobair—* The Well Field”; almost the only interest in the name is that there is neither well nor trace of one at present. “The Nursery,” a field. : The Gownach Well. Gambhnach = a yearling heifer; the eponymous bovine may be the legendary Glas. G. Gleann Caoin—“ Pleasant Glen.” Cnoicin na bPoll—* Little Hill of the Holes (Fox Earths).” Sruthan a Ghabha—‘ Little Stream of the Smith.” Paire a Raibe—* The Rape (crop) Field.” KILLaWILiyy, Cill a Mhuilinn—* Church of (near) the Mill.” Area, 540. Kilnemollan (Deps. 1652). The ancient mill, from proximity to which the ceall got its name, stood close to the site of the present Bride Bridge. The ceall itself is really in Deerpark townland, g.v. There was also a single large lios of which only a small are survives. Possibly, when he had got thus far, some mishap overtook the demolisher and put an untimely end to his labour. On O’Connor’s farm is a curious boulder, which, however, lacks a name. S8.DD. Poillin Bhriain—“Brian’s Little Pool”: this is a hole in the river. “The Well Ford”; an old crossing-place of the Bride. Kitt St. Anne, Cill Sant Anua—* Church of St. Anne.”’ Almost certainly there was an early Celtic church, called perhaps by name of the original founder, and for this the name of the Liblical saint was substituted in the early post-Invasion period: ¢f. Kill St. Laurence and Kill St. Nicholas,’ Co Waterford. Area, in two divisions, 844 4. On the townland are the remains of Castlelyons parish church, the ruins of Barrymore castle or mansion, and an early church site—probably the name-giving Ceall. The extensive remains of the mansion indicate a Tudor * Since I wrote the above Mr. R. A. Foley has suggested to me that the name is really Cill Sanctain, ‘‘ Church of S. Sanctan.” Powrr— Place-Numes and Antiquities of S.L. Cork. aly erection, rebuilding, or modification. The castle was destroyed by an acci- dental fire, July 22nd, 1771. It is said that the ravayes of the fire could have been easily stayed, but artizans, anticipating repairs, made no effort at salvage, and so the great building was gutted. Decay of the township followed as a result. Within thirty years from the fire not a shrub or tree remained of the once well-kept grounds and beautiful gardens. Of the Barrymores of Castlelyons almost the only one who has left a memory is Earl James, who earned the reputation of an able general in the Peninsular War (1807-14). Subsequently he fell into disgrace, and lost his command, Returning to Ireland, he tyrannized over his tenantry and retainers, at the same time that he lavished hospitality on his friends. His bust, in white marble, adorns the family mausoleum at Castlelyons. Of the ancient parish church of Castlelyons but little survives; this little, however, is sufficient to indicate that the church was of rather unusual size and architectural importance. In fact, the remains suggest rather a con- ventual than an Irish parochial church. An ivy-clad, square-planned tower, twenty-two feet to the side interiorly, and of which only the north- west angle stands entire, rises to a considerable height. This appears to have marked junction of nave and chancel. Practically nothing remains of the nave, which was 22 feet 6 inches in internal width; its site is occupied by the present Protestant church. A mausoleum of the Barrymore family, in hideous taste, has been erected on what was the east gable and end of chancel, and it is scarcely uncharitable to assume that the materials of the chancel were used up in erection of the monstrosity. At any rate, both north and south side walls of the chancel have disappeared, though their foundations are traceable. A high, pointed chancel-arch, forming the eastern support of the tower, still stands; it is 12 feet 6 inches wide, and its walls—indeed all the ground-floor walls of the tower—are 3 feet 10 inches thick. From presence of a second similar arch, at right angles to the chancel arch and forming the northern support of the tower, we infer—(1) position of the tower itself, ie. ab junction of nave and chancel; and (2) that the church had either a north transept or a north aisle. The early (apparently four- teenth-century) Gothic window in the east gable of the modern church was, I have little doubt, transferred to its present position from (probably the chancel of) the older church. ‘his handsome window is 48 inches wide, and is divided by two upright mullions into three lights. The well-kept cemetery, about three-fourths of an acre in extent, and crowded with tombstones, possesses no inscription or monument of special interest—at least, a fairly careful search, made in a downpour of rain, did not reveal anything such. Probably the oldest record is that borne by a standing stone at east side of [31*] 218 Proceedings of the Royal Irish Academy. the Barrymore vault, whither, presumably, it had been removed at building of the mausoleum :— “Here Lyeth the Body of John Newton Who departed this Life the 26th of July 1748 aged 27 Y*.” S.DD. Loughapreacane cross-roads (O.M.). Loch a Phréaciin. See under Corrin, antea. Aghnageragh (O.M.). ‘Ath na gCaorach—“Sheep Ford”; perhaps sheep were washed (or drowned) at this place. An Strapa—‘ The Stile,” a field. “Thom’s Hole,” a river hole, in which the man from whom it takes its name was drowned. “Flower Hill,” a subdivision—now almost forgotten—containing about forty acres. “The Long Walk,” a field. “The Ceall”’; on Barry’s farm. Piire a Bheistin—‘ Field of the Little Vest.” A battle is said to have been fought here. This was, doubtless, the battle of 1645 (May 10th) between the forces of Broghill and the Irish, in which the latter were defeated. “The Long Walk,” “The Castle Garden,” “The Great Meadow,” and “The Chapel Field” (at back of castle) = fields. Mouera, Mothara—* Ruins,” according to O’Donovan* (probably Stone ' Broghill, reporting to the English Parliament, thus describes the fight : ‘‘I ordered Major Paisley to keep the road with eighty horse in four ranks at the side of the hill which at some distance from the enemy looked like four battalions, with order when he had discharged his carbines to fly and rally in my rear. And having told my men that I would fight, and by God's blessing beat the enemy, I drew them to another piece of ground } mile further from the enemy who boldly came up, and having routed Paisley, pursued in disorder tili they perceived the rest of the horse in eleven battalions which encountered them fiercely ; but 800 Irish musqueteers coming up by the side of a ditch had like to ruin all if Captain Paisley with my own troop had not leaped over the ditch and defeated them with the loss of a hundred men on the spot. However, the Irish fought so well that one troop ran away to Castlemartyr with the news that all was lost; but the rest stood to it so well fhat we gained a noble victory, and if we had but 500 foot we might defeat their whole army, and had done it as it was but for an Irish sergeant with forty musqueteers, who, being posted in the wood, fired so often as though their whole foot were there. We did not lose one officer, and had only a few wounded. My horse was shot in the neck. The men had been twelve hours either marching, drawing, or fighting. Ridgeway, though drunk, killed nine that day with his own hand; his drunkenness was due to two tumblers of Rigley ale which he had from an Irish Sutler.” (Ms. in Lismore Castle, quoted in Ord. Survey Field Book.) * Field Book, Ord. Survey. Powrer—LPlace-Names and Antiquities of S.l’. Cork. 219 Fort in ruins). ‘he usual plural is Mothair. The local pronunciation is, however, Machaire—“ Plain (or battle field).” Area, 904 A. The sole object of antiquity is a pillar-stone of the type which I have designated Castlelyons. The present specimen, which stands on Sweeny’s farm, measures 5 feet by 5 feet by 2 feet. S.DD. Paire na Carraige—“