Newton flows for elliptic functions IV

An elliptic Newton flow is a dynamical system that can be interpreted as a continuous version of Newton’s iteration method for finding the zeros of an elliptic function f . Previous work focusses on structurally stable flows (i.e., the phase portraits are topologically invariant under perturbations of the poles and zeros for f), including a classification / representation result for such flows in terms of Newton graphs (i.e., cellularly embedded toroidal graphs fulfilling certain combinatorial properties). The present paper deals with non-structurally stable elliptic Newton flows determined by pseudo Newton graphs (i.e., cellularly embedded toroidal graphs, either generated by a Newton graph, or the so called nuclear Newton graph, exhibiting only one vertex and two edges). Our study results into a deeper insight in the creation of structurally stable Newton flows and the bifurcation of non-structurally stable Newton flows. Subject classification: 05C75, 33E05, 34D30, 37C70, 49M15.


Elliptic Newton flows; structural stability
The results in the following four subsections, can all be found in our paper [2].

Planar and toroidal elliptic Newton flows
Let f be an elliptic (i.e., meromorphic, doubly periodic) function of order r( 2) on the complex plane C with (ω 1 , ω 2 ), Im ω2 ω1 > 0, as basic periods spanning a lattice Λ(= Λ ω1, ω2 ). The planar elliptic Newton flow N (f ) is a C 1 -vector field on C, defined as a desingularized version 1 of the planar dynamical system, N (f ), given by: On a non-singular, oriented N (f )-trajectory z(t) we have: -arg (f ) =constant, and |f (z(t))| is a strictly decreasing function on t.
So that the N (f )-equilibria are of the form: -a stable star node (attractor); in the case of a zero for f , or -an unstable star node (repellor); in the case of a pole for f , or -a saddle; in the case of a critical point for f (i.e., f vanishes, but f not).
For an (un)stable node the (outgoing) incoming trajectories intersect under a nonvanishing angle ∆ k , where ∆ stands for the difference of the argf -values on these trajectories, and k for the multiplicity of the corresponding (pole) zero. The saddle in the case of a simple critical point (i.e., f does not vanish) is orthogonal and the two unstable (stable) separatrices constitute the "local" unstable (stable) manifold at this saddle.

The topology τ0
It is not difficult to see that the functions f , and also the corresponding toroidal Newton flows, can be represented by the set of all ordered pairs ({[a 1 ],· · ·, [a r ]}, {[b 1 ],· · ·, [b r ]}) of congruency classes mod Λ (with a i , b i ∈ P, i = 1, . . . , r,) that fulfil (3).
This representation space can be endowed with a topology, say τ 0 , induced by the Euclidean topology on C, that is natural in the following sense: Given an elliptic function f of order r and ε > 0 sufficiently small, a τ 0 -neighborhood O of f exists such that for any g in O, the zeros (poles) for g are contained in ε-neighborhoods of the zeros (poles) for f . E r (Λ) is the set of all functions f of order r on T (Λ) and N r (Λ) the set of corresponding flows N (f ). By X(T ) we mean the set of all C 1 -vector fields on T , endowed with the C 1 -topology (cf. [6]).
The topology τ 0 on E r (Λ and the C 1 -topology on X(T ) are matched by:

Canonical forms of elliptic Newton flows
The flows N (f ) and N (g)) in N r (Λ) are called conjugate, denoted N (f ) ∼ N (g), if there is a homeomorphism from T onto itself mapping maximal trajectories of N (f ) onto those of N (g), thereby respecting the orientations of these trajectories.
Consequently, unless strictly necessary, we suppress the role of Λ and write: E r (Λ) = E r , T (Λ) = T and N r (Λ) = N r .
-A structurally stable N (f ) has precisely 2r different simple saddles (all orthogonal).
The main results obtained in [2] are: The following three subsubsections describe shortly the main results from our paper [3]. Hence, by duality, It follows that also G * (f ) is cellularly embedded and fufills the E-and A-property. 6 i.e., all zeros, poles and critical points for f are simple, and no critical points are connected by N (f )trajectories. 7 The graph G(f ) has no loops, basically because the zeros for f are simple. A connected multigraph G in T with r vertices, 2r edges and r faces is called a Newton graph (of order r) if this graph is cellularly embedded and moreover, the A-property and the E-property hold. It is proved that the dual G * of a Newton graph G is also Newtonian (of order r). The anti-clockwise (clockwise) permutations on the edges of G at its vertices endow a clockwise (anti-clockwise) orientation of the G-faces and, successively, an anti-clockwise (clockwise) orientation of the G * -faces, compare Fig.2).

Newton graphs
-Apparently G(f ) and G * (f ) are Newton graphs.
The main results obtained in [3] are: -If N (f ) and N (g) are structurally stable and of the same order, then: -given a clockwise oriented Newton graph G of order r, there exists a structurally stable Newton flow N (f G ) such that G(f G ) ∼ G and thus: (H is another clockwise oriented Newton graph) Here the symbol ∼ between flows stands for conjugacy, and between graphs for equivalency (i.e. an orientation preserving isomorphism).

Characteristics for the A-property and the E-property
Let G be a cellularly embedded graph of order r in T , not necessarily fulfilling the A-property. There is a simple criterion available for G to fulfil the A-property. In order to formulate this criterion, we denote the vertices and faces of G by v i and F j respectively, i, j = 1, · · · , r. Now, let J be a subset of {1, · · · , r} and denote the subgraph of G, generated by all vertices and edges incident with the faces F j , j ∈ J, by G(J).The set of all vertices in G(J) is denoted by V (G(J)). Then: G has the A-property ⇔ |J| < |V (G(J))| for all J, ∅ = J {1,· · ·, r}, [Hall condition] where | · | stands as usual for cardinality. As a by-product we have: -Under the A-property, the set of exterior G(J)-vertices (i.e., vertices in G(J) that are also adjacent to G-faces, but not in G(J)), is non-empty.
-Let G be a cellularly embedded graph of order r in T , not necessarily fulfilling the E-property. We consider the rotation system Π for G: where the local rotation system π v at v is the cyclic permutation of the edges incident with v such that π v (e) is the successor of e in the anti-clockwise ordering around v. Then, the boundaries of the faces of G are formally described by Π-walks as: If e(= (v v )) stands for an edge, with end vertices v and v , we define a Π-walk (facial walk), say w, on G as follows: [face traversal procedure] Consider an edge e 1 = (v 1 v 2 ) and the closed walk w = v 1 e 1 v 2 e 2 v 3 · · · v k e k v 1 , which is determined by the requirement that, for i = 1, · · · , , we have π vi+1 (e i ) = e i+1 , where e +1 = e 1 and is minimal. 8 Each edge occurs either once in two different Π-walks, or twice (with opposite orientations) in only one Π-walk. G has the E-property iff the first possibility holds for all Π-walks. The dual G * admits a loop 9 iff the second possibility occurs at least in one of the Π-walks. The following observation will be referred to in the sequel: Under the E-property for G, each G-edge is adjacent to different faces; in fact, any G-edge, say e, determines precisely one G * -edge e * (and vice versa) so that there are 2r intersections s = (e, e * ) of G-and G * -edges. We consider an abstract graph with these pairs s, together with the G-and G * -vertices, as vertices, two of the vertices of this abstract graph being connected iff they are incident with the same G-edge or G * -edge. This graph admits a cellular embedding in T (cf. [3]), which will be referred to as to the "distinguished" graph G ∧ G * with as faces the so-called canonical regions (compare Fig.3). Following [10], G ∧ G * determines a C 1 -structurally stable flow X(G) on T . In [3] we proved that, if the A-property holds as well, X(G) is topologically equivalent with a structurally stable elliptic Newton flow of order r.

The Newton graphs of order r, r = 2, 3
Following [4], we present the lists of all -up to duality and conjugacy -Newton graphs of order r, r = 2, 3.

Pseudo Newton graphs
Throughout this section, let G r be a Newton graph of order r.
Due to the E-property, we know that an arbitrary edge of G r is contained in precisely two different faces. If we delete such an edge from G r and merge the involved faces F 1 , F 2 into a new face, say F 1,2 , we obtain a toroidal connected multigraph (again cellularly embedded) with r vertices, 2r − 1 edges and r − 1 faces: F 1,2 , F 3 , · · · , F r . If r = 2, then this graph has only one face. If r > 2, put J = {1, 2}, thus ∅ = J {1,· · ·, r}. Then, we know, by the A-property (cf. Subsubsection 1.2.3) , that the set Ext(G(J)) of exterior G(J)-vertices is non-empty. Let v ∈ Ext(G(J)), thus v ∈ ∂F 1,2 . Hence, v is incident with an edge, adjacent only to one of the faces F 1 , F 2 . Delete this edge and obtain the "merged face" F 1,2,3 . If r = 3, the result is a graph with only one face.
If r > 3, put J = {1, 2, 3}. By the same reasoning as used in the case r = 3, it can be shown that ∂F 1,2,3 contains an edge belonging to another face than F 1 , F 2 or F 3 , say F 4 . Delete this edge and obtain the "merged face" F 1,2,3,4 . And so on. In this way, we obtainin r−1 steps -a connected cellularly embedded multigraph, sayǦ r , with r vertices, r + 1 edges and only one face.
Obviously,Ǧ r contains vertices of degree 2. Let us assume that there exists a vertex foř G r , say v, with deg(v)=1. If we delete this vertex fromǦ r , together with the edge incident with v, we obtain a graph with (r − 1) vertices, r edges and one face. If this graph contains also a vertex of degree 1, we proceed successively. The process stops after L steps, resulting into a (connected, cellularly embedded) multigraph, sayĜ ρ . This graph admits ρ = r −L vertices (each of degree 2), ρ + 1 edges and one face.  Fig.3, it follows that G 2 is unique (up to equivalency). From the forthcoming Corollary 2.2 it follows that alsoǦ 2 is unique. However, a grapȟ G r , r > 2, is not uniquely determined by G r , as will be clear from  (b) There is a closed, clockwise oriented facial walk, say w, of length 2(ρ + 1) such that, traversing w, each vertex v shows up precisely deg(v) times. Moreover, w is divided into subwalks W 1 , W 2 , · · · , connecting vertices of degree > 2 that, apart from these begin-and endpoints, contain only-if any-vertices of degree 2.
(c 1 ) If e 1 e 2 · · · e s is a walk of type W i , then also W −1 , where e and e −1 stand for the same edge, but with opposite orientation.
Note that one uses here also thatĜ ρ has ρ vertices and all these vertices have degree 2. Thus, either The geometrical dual (Ĝ ρ ) * ofĜ ρ has only one vertex. So, all edges of (Ĝ ρ ) * are loops. Hence, in the facial walk w ofĜ ρ , each edge shows up precisely twice (with opposite orientation) cf. Subsubsection 1.2.3. Thus w has length 2(ρ + 1). By the face traversal procedure, each facial sector ofĜ ρ is encountered once and -at a vertex v-there are deg(v) many of such sectors. Application of (a 1 ) and (a 2 ) yields the second part of the assertion.
ad (c 1 ): Let e 1 ve 2 be a subwalk of w with deg(v) = 2. Both e −1 1 and e −1 2 occur precisely once in w, and e −1 2 ve −1 1 is a subwalk of w. ad (c 2 ): If the subwalk W = e 1 e 2 · · · e s and its inverse are consecutive, then -by the face traversal procedure-e −1 1 v 1 e 1 , or e s v s e −1 s are subwalks of the facial walk w. In the first case, v 1 is both begin-and endpoint of e 1 ; in the second case, v s is both begin-and endpoint of e s . This would imply thatĜ ρ has a loop which is excluded by construction of this graph. ad (c 3 ): Note that, traversing w once, each of the two vertices of degree 3 is encountered thrice. Suppose that the begin en and points of one of the subwalks W 1 , W 2 , · · · , say W 1 , coincide. Then, this also holds for the subwalk W −1 1 being-by (c 2 )-not adjacent to W 1 . So, traversing w once, this common begin/endpoint is encountered at least four times; in contradiction with our assumption. The assertion is an easy consequence of (a 1 ) and (c 2 ). ad (c 4 ): Traversing w once, the vertex of degree 4 is encountered four times. In view of (a 2 ) and (c 2 ), we find four closed subwalks of w namely: W 1 , W −1 1 (not adjacent) and W 2 , W −1 2 (not adjacent). Finally we note that each of these subwalks must contain at least one vertex of degree 2 (sinceĜ ρ has no loops).
An analysis of its rotation system learns thatĜ ρ is determined by its facial walk w, and thus also by the subwalks W 1 W 2 W 3 (in Case a 1 ) or W 1 W 2 (in Case a 2 ). In fact, only the length of the subwalks W i matters.
Corollary 2.2. The graphsĜ 2 andĜ 3 can be described as follows: • By Lemma 2.1 it follows thatĜ 2 does not have a vertex of degree 4. So,Ĝ 2 is of the form as depicted in Fig.6-a 1 , where each subwalk W i admits only one edge. Hence, there is-up to equivalency-only one possibility forĜ 2 . Compare also Fig.3.
• It is easily verified that -in Fig.7-each graph (on solid and dotted edges) is a Newton graph (cf. Subsubsection 1.2.3 ). Hence, in case ρ = 3, both alternatives in Lemma 2.1-(a) occur. An analysis of their rotation systems learns that the three graphs with only solid edges in Fig.7-(i) − (iii) are equivalent, but not equivalent with the graph on solid edges in Fig.7-(iv). In a similar way it can be proved that the graphs in Fig.7 expose all possibilities(up to equivalency) forĜ 3 .

Definition 2.3. Pseudo Newton graphs
Cellularly embedded toroidal graphs, obtained from G r by deleting edges and vertices in the way as described above, are called pseudo Newton graphs (of order r). Figure 6: The graphsĜ ρ .
Apparently, a pseudo Newton graph is not a Newton graph by itself. Replacing (in the inverse order) intoĜ ρ , ρ = r − L, the edges and vertices that we have deleted from G r , we regain subsequentlyǦ ρ and G r . A pseudo Newton graph of order r has either one face with angles summing up to the number (ρ ) of its vertices and (ρ + 1) edges (2 ρ ρ r) or one face with angles summing up to r , r − r faces with angles summing up to 1 and altogether 2r − r + 1 edges (2 r, 1 < r < r).
Remark 2.4. If we delete fromĜ ρ an arbitrary edge, the resulting graph remains connected, but the alternating sum of vertices, edges and face equals +1. Thus one obtains a graph that is not cellularly embedded. Apparently a Nuclear Newton graph is connected and admits one face and two loops. In particular such a graph has a trivial rotation system. Hence, all nuclear Newton graphs are topologically equivalent and since they expose the same structure as the pseudo Newton graphsĜ ρ , they will be denoted byĜ 1 . Note that a nuclear Newton graph fulfils the Aproperty (but certainly not the E-property). Consequently, a graph of the typeĜ 1 is neither a Newton graph, nor equivalent with a graph G(f ), f ∈Ě r . Nevertheless, nuclear Newton graphs will play an important role because, in a certain sense, they "generate" certain structurally stable Newton flows. This will be explained in the sequel.

Nuclear elliptic Newton flow
Throughout this section, let f be an elliptic function with -viewed to as to a function on T = T (Λ(ω 1 , ω 2 )) -only one zero and one pole, both of order r, r 2. Our aim is to derive the result on the corresponding (so-called nuclear) Newton flow N (f ) that was already announced in [2], Remark 5.8. To be more precise: " All nuclear Newton flows -of any order r-are conjugate , in particular each of them has precisely two saddles (simple) and there are no saddle connections".
When studying -up to conjugacy-the flow N (f ), we may assume (cf. Subsubsection 1.1.3) that ω 1 = 1, ω 2 = i, thus Λ = Λ 1,i . In particular, the period pair (1, i) is reduced. We represent f (and thus N (f )), by the Λ-classes [a], [b] , where a resp. b stands for the zero, resp. pole, for f , situated in the period parallelogram P (= P 1,i ). Due to (2), (3) we have: We may assume that a, b are not on the boundary ∂P of P . Since the period pair (1, i) is reduced, the images under f of the P -sides γ 1 and γ 2 are closed Jordan curves (use the explicit formula for λ 0 as presented in Footnote 3). From this, we find that the winding numbers η(f (γ 1 )) and η(f (γ 2 )) can -a priori-only take the values -1, 0 or +1. The combination (η(f (γ 1 )), η(f (γ 2 ))) = (0, 0) is impossible (because a = b). The remaining combinations lead, for each value of r = 2, 3, · · · , to eight different values for b each of which giving rise, together with a, to eight pairs of classes mod Λ that fulfil (2), determining flows in N r (Λ), compare Fig. 8, where we assumed -under a suitable translation of P -that a = 0. Note that the derivative f is elliptic of order r + 1. Since there is on P only one zero for f (of order r), the function f has two critical points, i.e., saddles for N (f ), counted by multiplicity.
The eight pairs (a, b) that possibly determine a nuclear Newton flow are subdivided into two classes, each containing four configurations (a, b): (see Fig.9) Class 1: a = 0, b on a side of the period square P . Class 2: a = 0, b on a diagonal of the period square P , but not on ∂P .
Apparently, two nuclear Newton flows represented by configurations in the same class are related by a unimodular transformation on the period pair (1, i), and are thus conjugate, see Subsubsection 1.1.3. So, it is enough to study nuclear Newton flows, possibly represented by (0, 1 r ) or (0, 1+i r ).
The configuration (a, b), a = 0, b = 1 r : The line 1 between 0 and 1, and the line 2 between 1 i and 1 + 1 i , are axes of mirror symmetry with respect to this configuration (cf. Fig.10-(i)). By the aid of this symmetry and using the double periodicity of the supposed flow , it is easily proved that this configuration can not give rise to a desired nuclear Newton flow.
The configuration (a, b), a = 0, b = 1+i r : The line between 0 and 1+i 2 is an axis of mirror symmetry with respect to this configuration (cf. Fig.10-(ii)). So the two saddles of the possible nuclear Newton flow are situated either on the diagonal of P through 1+i r , or not on this diagonal but symmetric with respect to 1 . The first possibility can be ruled out (by the aid of the symmetry w.r.t. and using he double periodicity of the supposed flow). So it remains to analyze the second possibility. (Note that only in the case where r = 2, also the second diagonal of P yields an axis of mirror symmetry). We focus on Fig.11, where the only relevant configuration determining a (planar) flow N (f ), is depicted. By symmetry, the -segments between 0 and 1+i r , and between 1+i r and (1+i) are N (f )-trajectories connecting the pole 1+i r , with the zeros 0 and (1+i). Since on the N (f )-trajectories the arg(f ) values are constant, we may arrange the argument function on C such that on the segment between 1+i r and 1 + i we have arg(f )= 0. We put argf (σ 1 )= α, thus 0 < α < 1 and arg f (σ 2 )= −α. Note that at the zero / pole for f , each value of arg(f ) appears r times on equally distributed incoming (outgoing) N (f )-trajectories. By the aid of this observation, together with the symmetry and periodicity of f , we find out that the phase portrait of N (f ) is as depicted in Fig.11, where box stands for the (constant values of arg(f ) on the unstable manifolds of N (f ). In particular, there are no saddle connections.   From Fig.11 it is evident that H r (f ) is a cellularly embedded pseudo graph (loops and multiple edges permitted). This graph is referred to as to the nuclear Newton graph for N (f ). By Lemma 3.2, the graphs H r (f ), are -up to equivalency -unique, and will be denoted by H r (compare the comment on Definition 2.5).
If a, b (both in P (= P 1,i )) are of Class 2 (i.e., the configuration (a, b) determines a nuclear flow with a and b as zero resp. pole of order r), we introduce the doubly periodic functions: where the summation takes place over all points in lattice Λ(= Λ 1,i ). We define the planar flow N (f ) by: Lemma 3.4. The flow N (f ) is smooth on C and exhibits the same phase portrait as N (f ), but, its attractors (at zeros for f ) and its repellors (at the poles for f ) are all generic, i.e. of the hyperbolic type.  Proof. Let N (f ) be arbitrary. Because all its equilibria are generic and there are no saddle connections, this flow is C 1 -structurally stable. The embedded graph H r (= H r (f )), together with its geometrical dual H r (f ) * , forms the so-called distinguished graph that determines -up to an orientation preserving homeomorphism -the phase portrait of N (f ) (cf. [10], [11] and Subsubsection 1.2.3). This distinghuished graph is extremely simple, giving rise to only four distinghuished sets (see Fig.11). This holds for any flow of the type N (f ). Now, application of Peixoto's classification theorem for C 1 -structurally stable flows on T yields the assertion.
Thus (p 1 , p 2 ) and (q 1 , q 2 ) are co-prime, and Our aim is to describe H(f ω1,ω2 ) as a graph on the canonical torus T (= T 1,i ). In view of Lemma 3.2, the two edges of H(f ω1,ω2 ) are closed Jordan curves on T , corresponding to the unstable manifolds of N (f ω1,ω2 )) at the two critical points for f that are situated in the period parallelogram P ω1,ω2 . These unstable manifolds connect a(= 0) with p 1 + p 2 i, and q 1 + q 2 i respectively. Hence, one of the H(f ω1,ω2 )-edges wraps p 1 -times around T in the direction of the period 1 and p 2 -times around T in the direction of the period i, whereas the other edge wraps q 1 -times around this torus in the 1-direction respectively q 2 -times in the i-direction. See also Fig.13, where we have chosen for f the Weierstrass ℘-function (lemniscate case), i.e. r = 2, a = 0 and ω 1 = 3 + i, ω 2 = 2 + i. Compare also Fig.12, case r = 2. Figure 13: The nuclear Newton graph H(℘ 3+i,2+i ) on the torus T (= T 1,i ).

The bifurcation & creation of elliptic Newton flows
In this section we discuss the connection between pseudo Newton graphs and Newton flows. In order not to blow up the size of our study, we focus -after a brief introduction -on the cases r = 2, 3. However, even from these simplest cases we get some flavor of what we may expect when dealing with a more general approach.
We consider functions g ∈ E r with r simple zeros and only one pole (of order r); such functions exist, compare Subsubsection 1.1.1. The set of all these functions is denoted by Figure 14: The three different pseudo Newton graphsǦ 3 ,Ĝ 3 . E 1 r and will be endowed with the relative topology induced by the topology τ 0 on E r . Since the derivative g of g is elliptic of order r + 1, the zeros for g being simple, there are r + 1 critical points for g (counted by multiplicity).
We consider the set N 1 r of all toroidal Newton flows N (g). Such a flow is C 1 -structurally stable (thus also τ 0 -structurally stable) if and only if: (cf. subsection 1. 2. There are no "saddle connections".
3. The repellor at the pole for g is generic.
In general none of these conditions is fulfilled. We overcome this complication as follows: ad 1. Under suitably chosen -but arbitrarily small -perturbations of the zeros and poles of g, thereby preserving their multiplicities, N (g) turns into a Newton flow with only simple (thus r + 1) saddles (cf. [2], Lemma 5.7, case A = r, B = 1). ad 2. Possible saddle connections can be broken by adding to g a suitably chosen, but arbitrarily small constant (cf. [2], proof of Theorem 5.6 (2)). ad 3. With the aid of a suitably chosen additional damping factor to N (g), the pole of g may be viewed to as generic for the resulting flow; compare the proof of Lemma 3.4. (Note that the simple zeros for g yield already generic equilibriae).
This opens the possibility to adapt g and N (g) in such a way that for almost all functions g the flow N (g) is structurally stable (see Subsubsection 1.1.4, and Theorem 5.6 in [2]). More formally: The set E 1 r of functions g in E 1 r , with N (g) structurally stable, is τ 0 -open and -dense in E 1 r .
From now on, we assume that N (g) is structurally stable and define the multi graph G r (g) on T as follows: -Vertices: r zeros for g (i.e., stable star nodes for N (g)).
-Edges: r + 1 unstable manifolds at the critical points for g (orthogonal saddles for N (g)).
-Face: the basin of repulsion of the unstable star node at the pole for g. Note that G r (g) has no loops (since the zeros for g are simple).
It is easily seen that G r (g) is cellularly embedded (cf. [3], proof of Lemma 2.9).
Because G r (g) has only one face, the geometrical dual G r (g) * admits merely loops and the Π-walk for the G r (g)-face consists of 2(r + 1) edges, each occurring twice, be it with opposite orientation; here the orientation on the Π-walk is induced by the anti-clockwise orientation on the embedded G r (g) * -edges at the pole for g. In the case where G r (g) admits a vertex of degree 1, we delete this vertex together with the adjacent edge, resulting into a cellularly embedded graph on r − 1 vertices, r edges and only one face. If this graph has a vertex of degree 1, we repeat the procedure, and so on. The process stops after L (< r − 1) steps, resulting into a connected, cellularly embedded muligraph of the typê G ρ , ρ = r − L, 2 ρ r. Now, we raise the question whether the graphs obtained in this way are indeed pseudo Newtonian, i.e., do they originate from a Newton graph? And even so, can all pseudo Newton graphs be represented by elliptic Newton flows?
In the sequel we give an (affirmative) answer to these questions only in the cases r = 2 and r = 3.
Lemma 4.1. If r = 2 or 3, then the graph G r (g), g ∈ E 1 r , is a pseudo Newton graphǦ r .
Proof. Firstly, note that the proof of Lemma 2.1 does not rely on the fact thatĜ ρ originates from Newton graphs, but merely on the cellularity ofĜ ρ in combination with the property that #{edges} = 1 + #{vertices}. Case r = 2: By Corollary 2.2 of Lemma 2.1 we know:Ĝ 2 (=Ǧ 2 ) is unique (up to equivalency). So, G 2 (g) has the same topological type asĜ 2 and originates from a Newton graph (compare Fig.3 and Fig.6a 1 , where all subwalks W i admit only one edge). Case r = 3: If G 3 (g) has a vertex v 1 of degree 1, the graph obtained by deleting v 1 together with the adjacent edge c is a cellularly embedded graph in T with two vertices and three edges and mustĜ 2 . So G 3 (g) is of the form Fig.14(a). If G 3 (g) has no vertex of degree 1, this graph is of typeǦ 3 , and thus -by Corollary 2.2either of the form as depicted in Fig.14 (b), or Fig.14 (c).
So, we find that G 3 (g) takes, a priori, the three possible forms in Fig.14, where the values of degree v i discriminate between these possibilities. Recall that these three graphs originate from Newton graphs.
The reasoning in the above Case r = 3 does not imply that each of the graphs in Fig.14 can be realized by a Newton flow. So we need: Lemma 4.2. If r = 2 or 3, then each pseudo Newton graph of the typeǦ r orĜ r can be represented as G r (g), g ∈ E 1 r .
Proof. r = 2 : Follows from Lemma 4.1. r = 3 : In Fig.16 we consider the local phase portrait of N (f ) around the zero v(= 1 + i) for f . Compare Fig.11 and note that the zeros for f are star nodes for N (f ). Since α+γ = 1 r −α and α + β + γ = 1 2 we have β > 1 2r so that the angle β + β spans an arc greater than 1 r (= 1 3 ). Now the idea is: To split off from the 3 rd order zero v for f a simple zero (v 1 ) "Step 1", and thereupon, to split up the remaining double zero (v 1 ) into two simple ones (v 2 , v 3 ) "Step 2", in such a way that by an appropriate strategy, the resulting functions give rise to Newton flows with associated graphs, determining each of the three possible types in Fig.14. Ad Step 1 : We perturb the original function f into an elliptic function g with one simple (v 1 ) and one double (v 1 ) zero (close to each other), and one third order pole w 1 (thus close 11 11 Use property (2). and v 1 as attractors and w 1 as repellor. When v 1 tends to v 1 , the perturbed function g will tend to f , and thus the perturbed flow N (g) to N (f ), cf. Subsubsection 1.1.2. In particular, when the splitted zeros are sufficiently close to each other and the circle C 1 that encloses an open disk D 1 with center v 1 , is chosen sufficiently small, C 1 is a global boundary (cf. [7]) for the perturbed flow N (g). It follows that, apart from the equilibria v 1 and v 1 (both of Poincaré index 1) the flow N (g) exhibits on D 1 one other equilibrium (with index −1): a simple saddle, say c (cf. [5]). From this, it follows (cf. Subsubsection 1.1.1) that the phase portrait of N (g) around v 1 and v 1 is as sketched in Fig.15-(a), where the local basin of attraction for v 1 is shaded and intersects C 1 under an arc with length approximately 1 3 . On the (compact!) complement T \D 1 this flow has one repellor (w 1 ) and two saddles. The repellor may be considered as hyperbolic (by the suitably chosen damping factor, compare the proof of Lemma 3.4), whereas the saddles are distinct and thus simple (because N (f ) has two simple saddles, say σ 1 , σ 2 , depending continuously on v 1 and v 1 ). Hence, the restriction of N (g) to T \D 1 is ε-structurally stable (cf. [10]). So, we may conclude that, if v 1 (chosen sufficiently close to v 1 ) turns around v 1 , the phase portraits outside D 1 of the perturbed flows undergo a change that is negligible in the sense of the C 1 -topology. Therefore, we denote the equilibria of N (g) on T \D 1 by w 1 , σ 1 , σ 2 (i.e., without reference to v 1 ). We move v 1 around a small circle, centered at v 1 and focus on two positions (I, II) of v 1 , specified by the position of v 1 w.r.t. the symmetry axis . See Fig.16 in comparison with Fig.17, where we sketched some trajectories of the phase portraits of N (g) on D 1 .

Ad
Step 2 : We proceed as in Step 1. Splitting v 1 into v 2 and v 3 (sufficiently close to each other) yields a perturbed elliptic function h, and thus a perturbed flow N (h). Consider a circle C 2 , centered at the mid-point of v 2 and v 3 , that encloses an open disk D 2 containing these points. If we choose C 2 sufficiently small, it is a global boundary of N (h). Reasoning as in Step 1, we find out that N (h) has on D 2 two simple attractors (v 2 , v 3 ) and one simple saddle: d (close to the mid point of v 2 and v 3 ; compare Fig.15-(b)), where the local basin of attraction for v 3 is shaded and intersects C 2 under an arc with length approximately 1 2 . Moreover, as for N (g) in Step 1, the flow N (h) is ε-structurally stable outside D 2 . So, we may conclude that, if v 2 and v 3 turn (in diametrical position) around their mid-point, the phase portraits outside D 2 of the perturbed flows undergo a change that is negligible in the sense of C 1 -topology. Therefore, we denote the equilibria of N (h) on T \D 2 by v 1 , w 1 , c, σ 1 ,  for v 1 in the position of Fig.17-(II), we distinguish between two possibilities: Fig.18-IIa or Fig.18-IIb. Note that, with these choices of v 1 , v 2 , v 3 each of the obtained functions has three simple zeros and one triple pole. Moreover, the four saddles are simple and not connected, whereas the three zeros are simple as well. So the graph of the associated Newton flow is well defined and has only one face, four edges and three vertices. Recall that the various values of degree v i discriminate between the three possibilities for the graphs of typeǦ 3 ). Now inspection of Fig.18 yields the assertion.
Up till now, we paid attention to pseudo Newton graphs with only one face (i.e., of typě G orĜ.). If r = 2, these are the only possibilities. If r = 3, there are also pseudo Newton graphs (denoted by G) with two faces and angles summing up to 1 or 2. When the boundaries of any pair of the original G 3 -faces have a subwalk in common, these walks have length 1 or 2. (Use the A-property and compare Fig.4). So, when two G 3 -faces are merged, the resulting G-face admits either only vertices of degree 2 or one vertex of degree 1 12 . From now on, we focus on the Newton graphs as exposured in Fig.4 Figure 19: The graph G ∧ G * .
We consider the common refinement G ∧ G * of G and its dual G * . Following Peixoto [10], [11] we claim that G ∧ G * determines a C 1 -structurally stable toroidal flow X(G) with canonical regions as depicted 13 in Fig.20. As equilibria for X(G) we have: three stable and two unstable proper nodes (corresponding to the G-resp. G * -vertices) and five orthogonal saddles (corresponding to the pairs (e, e * ) of G-and G * -edges).
Argueing basically as in the proof of Theorem 4.1 from our paper [3] , it can be shown that X(G)) is equivalent with an elliptic Newton flow generated by a function on three simple zeros, one double and one simple pole and five simple critical points; compare 14 Fig.10. As an elliptc Newton flow, X(G) is not (τ 0 -)structurally stable, since G is not Newtonian, compare Subsubsection 1.2.2. However, by the aid of a suitably chosen damping factor, compare Lemma 3.4 (preambule), and within the class of all elliptic Newton flows generated by functions on three simple zeros, on one double and one simple pole and on five simple critical points, the flow X(G) is structurally stable w.r.t. the relative topology τ 0 .
Altogether, we find: Theorem 4.3. Any pseudo Newton graph of order r, r = 2, 3 represents an elliptic Newton flow. In particular, the 3 rd order nuclear Newton flow "creates" -by splitting up zeros/ poles ("bifurcation")-all, up to duality and topological equivalency, structurally stable elliptic Newton flows of order 3.