---

\ihg CASE NO A 538/92

IN THE SUPREME COURT OF SOUTH AFRICA

(APPELLATE DIVISION)

In the matter between:

JAN DANIëL ROUX Appellant

and
VAALPLAAS CALEDON (EDMS) BPK Respondent

CORAM: JOUBERT, KUMLEBEN, NIENABER, JJA et VAN COLLER, MAHOMED AJJA

DATE OF HEARING: 17 MARCH 1994

DATE OF JUDGMENT: 29 MARCH 1994

JUDGMENT

MAHOMED AJA:
2 The Appellant was the Defendant in an action brought in the Magistrate's Court for the district of Caledon by the Respondent as the Plaintiff. It was alleged by the Respondent that a dog or dogs owned by the Appellant had on four separate occasions, during the period 2nd July 1988 to 23rd August 1988 bitten a considerable number of livestock which was being kept on the property of the Respondent, thereby causing damages to the Respondent. The Magistrate granted judgment in favour of the Respondent as claimed in the summons and that judgment was upheld on appeal by the Cape Provincial Division of the Supreme Court.

The Appellant attacks both these judgments on the grounds that the Respondent failed to discharge the onus of proving, on a balance of probabilities, that it was indeed any dog of the Appellant which caused the damage to the Respondent's livestock alleged in the summons.

3
The first of the four occasions referred to took place on the 2nd July 1988. Mr J B E De Wet ("De Wet"), the chief manager of the Respondent at its farm "Vaalplaas", testified that seventy-six lambs and two ewes died and another twenty-two sheep were injured in consequence of having been bitten by a dog but neither he nor any other witness gave any direct testimony as to which dog or dogs were involved in this attack on the Respondent's livestock. De Wet was able to say, however, that this damage was indeed caused by a dog or dogs, because of his training and experience which included studies in livestock, registration of animals and the management of sheep farms.

The second attack was on the 29th July 1988. On this occasion seventeen sheep died and another six were injured in consequence of dog bites. De Wet actually saw a dog among the sheep in this instance. When this dog noticed De Wet, it jumped over the fence.

4
De Wet followed it to an area near the home of the
Appellant. The dog went to the yard next to the house of

the Appellant. De Wet then went to the Appellant's house

to ask whether it was not the Appellant's dog which was

involved. By this time the dog which had been followed

by De Wet, was already on the premises of the Appellant

who denied that this dog had been out of the yard at all.

De Wet observed, however, that the dog was wet and out of

breath, De Wet testified that the Appellant tried to

convince him how secure his premises were but De Wet

noticed a hole through the fencing through which the dog

could easily have gone to the camp where the Respondent's

sheep had been bitten. De Wet said that he pointed this

out to the Appellant.

The third attack took place on the 17th August

1988 when twenty-four lambs were killed and another

eleven sheep were injured. De Wet was able to say that

they were all attacked by a dog but he did not actually
5 witness any dog on the Respondent's premises because he was attending a sale of rams in Bloemfontein on that day.
The inference that the dog which attacked the Respondent's livestock on this occasion was a dog belonging to the Appellant, is sought to be drawn from the expert testimony of Mr G W Dyer ("Dyer") , augmented by some circumstantial evidence. Dyer who had considerable experience with animals and particularly with dogs which attacked sheep, testified that once a dog had started biting sheep at a particular place, it was highly probable that he would return to repeat such conduct again.

The fourth occasion when the Respondent's sheep were attacked was on the 23rd August 1988. Twenty-seven lambs and five sheep were killed and another fifteen sheep were injured on that occasion. De Wet testified that he actually saw the dog on this occasion among the
6 Respondent's sheep and that it was the same dog which he had previously followed to the house of the Appellant. He said that although it was dark he was able to identify the dog with the lights of his vehicle and he had no doubt whatever that the dog was indeed the same dog which he had seen on the second occasion. De Wet said that he jumped out of his vehicle to open the gate but the dog managed to run away. De Wet fired one shot but missed the dog. De Wet testified that he then again went to the house of the Appellant who admitted to him that one of his dogs was not there. De Wet decided to wait for the return of the dog. He also contacted another manager of the Appellant and certain other workers by radio and asked them to find the dog. They did so and killed a dog at camp No. 7. That dog was then brought to the premises of the Appellant and immediately identified by De Wet as the dog he had seen earlier at the camp of the Respondent. The Appellant admitted that it was indeed
7 his dog.
The onus was clearly on the Respondent to establish, on a balance of probabilities, the allegation made in the summons that the dog which killed or injured its livestock, indeed belonged to the Appellant. If that onus was discharged the Respondent was entitled to recover from the Appellant such damages as were proved to have been suffered by the Respondent in consequence of the conduct of the Appellant's dog. The Appellant's liability arises in consequence of its ownership of the dog. It was not necessary for the Respondent to aver or establish dolus or culpa [Vermaak v Du Plessis 1974 (4) SA 353 (0); Coetzee & Sons v Smit and Another 1955 (2) SA 553 (A) et 558; D. 9.1.1;].

The Magistrate before whom the trial was conducted concluded that it was indeed the Appellant's dog which had been responsible for killing or injuring the livestock of the Appellant on all four occasions and
8 that the Respondent had established the quantum of its damages arising from these incidents. In coming to that conclusion the Magistrate accepted the evidence of both De Wet and Dyer. He regarded both of these witnesses as satisfactory. The Appellant and his main supporting witness, Verneul were considered by the Magistrate to be weak and unreliable witnesses. The Cape Provincial Division of the Supreme Court dismissed the Appeal against the Judgment of the Magistrate and concluded that there were no sound reasons why it should interfere with the credibility findings made by the Magistrate. It also concluded that the evidence justified the inference that the Appellant's dog was indeed involved in the attack on the Respondent's sheep on all four occasions and that the Respondent had established the quantum of its damages. On the merits of the Respondent's claim against the Appellant it is necessary to determine two issues: a) Was the dog which De Wet observed among the sheep on
9 the second occasion (29th July 1988) and the fourth occasion (23rd August 1988), a dog belonging to the Appellant? b) If it was, does the evidence justify the inference that this dog was also involved in the attack on the Respondents sheep on the first occasion (2nd July 1988) and on the third occasion (17th August 1988)? The identification of the dog.

There is considerable circumstantial evidence which supports the conclusion that the dog which De Wet saw on the second occasion was indeed the Appellant's dog. He was able to identify the dog whilst it was among the sheep. He followed it to a point next to the home of the Appellant and shortly thereafter the same dog came to the premises of the Appellant. All this transpired at approximately eight o'clock in the morning under normal conditions of visibility . Although the Appellant said to De Wet that all his dogs were on his premises on this
10 occasion, the force of that assertion is substantially debilitated by two objective sources of evidence. The first was a hole in the fencing on the premises which clearly made it possible for the dog to take a convenient route to the camp of the Respondent at which De Wet says he first saw the dog among the sheep. The second objective evidence was that the dog was wet and panting. This was corroborative of De Wet's testimony that the dog had been running from the camp, through some water and into the area occupied by the Appellant and his neighbour. That area was less than a kilometre away from the camp.

The cumulative effect of this evidence, in my view, justified the conclusion of the Magistrate that the dog which arrived panting at the premises of the Appellant, was indeed the dog which De Wet had seen earlier among the sheep at the camp of the Respondent.

Counsel for the Appellant did not suggest that

11

the dog which De Wet saw arriving, wet and panting, at
the premises of the Appellant did not belong to the
Appellant. What he did suggest was that it might not
have been the same dog which De Wet had seen among the
sheep earlier. In support of this attack on De Wet's
identification of the dog, Counsel referred to De Wet's
description of the dog as a "Rottweiler" with a "rooi-
bruin" colour and he argued that this was contradicted by
Dyer who said that a Rottweiler could not be described as
"rooi-bruin". In my view there is no real substance in
this criticism. What De Wet actually said was that the
colour appeared to him to be "rooi-bruin" and that it was
a "Rottweiler type". De Wet was never dogmatic that the
dog was indeed a Rottweiler. He said that it "appeared"
to him that it was a "Rottweiler type". That description
could not, at the time, have appeared to the Appellant to
be so much in conflict with the description of any dog
owned by the Appellant as to justify the cross-

12 examination of De Wet on this issue. The Appellant, in fact, conceded that he did not instruct his attorney at the time to confront De Wet in cross-examination on that basis.

The exact colour of the dog which De Wet saw among the sheep on the second occasion did not play any material role in his identification of the dog. What was more important was that he had followed the dog to the very area in which the Appellant had his home and that when he saw the wet and panting dog on the premises of the Appellant, he had not the slightest difficulty in identifying the dog as being the dog that he had seen shortly before, at the camp of the Respondent.

Both the direct testimony of De Wet and the circumstantial evidence in support thereof, therefore, justified the conclusion of the Magistrate that it was indeed the Appellant's dog which was involved in the attack on the Respondent's livestock on the second
13 occasion. There are, in my view, no grounds on which, this Court, is legitimately entitled to interfere with the findings of the Magistrate in this regard.

De Wet also had no doubt that the same dog was involved in the attack on the Respondent's animals on the fourth occasion. The Magistrate who accepted that testimony was alive to the fact that De Wet' s identification of the dog on the fourth occasion took place at approximately six fifteen on the morning of the 23rd August when it was still dark but De Wet was assisted in that identification by the head-lights of his vehicle. These lights reflected directly on the dog and enabled De Wet instantly to recognise the dog as being the same dog which was involved in the attack on the Respondent's livestock on the second occasion. De Wet also had the further opportunity of seeing the dog from close quarters as it jumped before him. He just missed shooting him on the spot.

14
De Wet's a identification of the dog on the fourth occasion is also supported by other circumstantial evidence. When he arrived at the home of the Appellant, it was conceded by the Appellant that one of his dogs was missing and after the dog had been shot by one of the employees of the Respondent, its body was immediately identified by De Wet as being the body of the dog he had seen earlier on the Respondent's farm. The Appellant admitted there and then that the dog which had been killed was indeed his dog. Moreover, with the death of this dog, the attacks on the livestock of the Respondent ceased completely. No further attacks of whatever nature took place on the Respondent's livestock in the ensuing period of more than a year before the trial was heard.

In my view, there was therefore sufficient evidence to justify the Magistrate's conclusion that the Appellant's dog was indeed involved in the attack on the Respondent's sheep on the fourth occasion. The Cape
15 Division of the Supreme Court came to the same conclusion and I agree that there are no grounds on which the findings of the Magistrate in this regard can be disturbed.

The conclusion that the dog of the Appellant was involved in the attack of the Respondent's sheep on the second and fourth occasions makes it unnecessary for me to deal with the suggestion that there were perhaps other dogs involved in these attacks. Not only is there no evidence to support any such suggestion but even if there were other dogs which had acted in concert with the Appellant's dog to attack the sheep of the Respondent and even if these other dogs belonged to other owners, the Appellant would remain liable for the damages suffered by the Respondent in consequence of such attacks. [Nel v Halse 6 S C 275; Katz v Bloomfield Keith 1914 T P D 379; O'Callaghan N. 0. v Chaplin 1927 A D 310]

The conclusion that the Appellant's dog was

16 involved in the attacks on the animals of the Respondent on the second and the fourth occasions is an important factor, in determining the liability of the Appellant in respect of the attack on the Respondent's animals, on the third occasion. Only nineteen days separated the second occasion from the third occasion and only six days separated the third occasion from the fourth occasion. Regard being had to the evidence of Dyer that once a dog has started biting sheep at a particular place it tends to return to repeat such attacks, it seems to me to be probable that having attacked the Respondent's animals on the second occasion, the Appellant's dog did indeed return to repeat his conduct on the third occasion. That probability is strengthened by the finding that it was indeed the same dog which attacked the Respondent's animals on the fourth occasion and the evidence that these attacks on the Respondent's animals ceased after the dog was shot and killed on the fourth occasion. It
17 is, of course, possible that although the Appellant's dog attacked the Respondent's animals on the second and the fourth occasion, some other dog and not the Appellant's dog was involved in the attack on the third occasion but is it probable? I think not. If there were any such other dogs why did the attack on the Respondent's animals cease after the dog belonging to the Appellant had been killed? The only possible explanation tendered for this, was a casual suggestion in the evidence on behalf of the Appellant to the effect that the killing of the Appellant's dog might have so frightened all other dog owners as to discipline them into taking effective steps to confine their dogs, thus frustrating the ambitions of those dogs which might have been involved in the attack on the Respondent's animals on the third occasion. Neither the Magistrate nor the Cape Provincial Division of the Supreme Court was very impressed with this suggestion. I agree. There are in my view no grounds

18
for interfering with the Magistrate's judgment with respect to the Appellant's liability for the damages caused to the Respondent during the attack on its animals on the third occasion.

Does the evidence justify the conclusion that the Appellant's dog was also responsible for the attack on the Respondent's animals on the first occasion? This is a more difficult question. The finding that the same dog was involved in the attacks on the Respondent's animals on the second, third and fourth occasions supports the conclusion that the same dog must have been involved in the first attack and to some extent that inference is strengthened by the fact that when the dog was killed on the fourth occasion the attacks on the Respondent's animals ceased. But in the assessment of probabilities these factors must be weighed against other considerations.

The inference that the Appellant's dog was

19 involved in the attack on the Respondent's animals on the third occasion is supported by the finding that the same dog was involved in a preceding attack on the second occasion but the conclusion that the Appellant's dog was also involved on the first occasion is not supported by any such preceding conduct. The evidence in this regard is equally consistent with the hypothesis that the second occasion constituted the very first escapade of the Appellant's dog in attacking animals on the property of the Respondent. That hypothesis is also supported by the difference in the quality and the intensity of the attacks. On the first occasion seventy-six lambs and two ewes were killed and another another twenty-two sheep were injured. This was a massive attack and substantially different from the second attack when seventeen sheep were killed and six were injured, or the third attack when twenty-four lambs were killed and eleven sheep were injured, or the fourth attack when

20
twenty-seven lambs and five sheep were killed and another fifteen sheep were injured. The differences are sufficiently significant to justify very considerable doubt as to whether the same dog was involved in all the attacks or whether the first attack was made by a different pack of dogs not Involving the Appellant's dog. (The evidence disclosed that there were other dogs in the vicinity which could have had access to the Respondent's premises and attacked its sheep). The period between the first and second attack is also not so short as to justify the inference that the dog involved in the attack on the first occasion must have returned on the second occasion because an interval of some twenty-seven days separated the two events.

It could be argued that there were no other dogs involved in the attack on the Respondent's animals on the first occasion, because on Dyer's evidence they could have been expected to return and repeat those
21 attacks. I do not think, however, that this consideration is sufficient to establish a balance of probabilities in favour of the case of the Respondent. These other dogs might have found other pastures or lost interest in the animals of the Respondent for a variety of reasons. Dyer's testimony was not that this never happens. Indeed, there had been prior attacks on the animals of the Respondent some three months before the second of July 1988 which was the date of the first occasion referred to in the summons. Those previous attacks ceased at some time and it is equally possible that the dog or dogs which attacked the Respondent's animals on the first occasion ceased their attacks on this farm for some or other reason. They may have been better secured within the premises of their owners or they may have found more convenient targets for attack or they may have been even destroyed.

The case for the Respondent in respect of the

22

second and the fourth occasion rested on the direct evidence of De Wet supported by significant circumstantial evidence. Its case in respect of the third occasion was supported by the finding that the Appellant's dog was involved in an attack on the second occasion and by circumstantial evidence. Its case in respect of the attack on the first occasion is, however, different: It is not supported by any direct evidence and the circumstantial evidence is not sufficiently compelling to enable the Respondent to contend that it has discharged its onus of proof on a balance of probabilities. In the result the finding of the Trial Court that the Appellant is liable for the damage suffered by the Respondent in consequence of the attack on its animals on the first occasion, must be set aside.

In support the Respondent's claim for damages De Wet gave evidence about the number of animals which had been killed and injured on the farm of the Respondent

23
in consequence of the acts of the dog belonging to the Appellant. Evidence was also led about the value of these animals and the cost of medication. There was, therefore, prima facie evidence in support of the Respondent's claim for damages in the summons. This evidence was accepted by the Magistrate. The Appellant led no evidence to contradict the Respondent's case on the quantification of damages. An attack on that quantification both in the Magistrates' Court and cm appeal before the Cape Provincial Division of the Supreme Court was in my view correctly rejected.

Although that attack was repeated in the Heads of Argument filed on behalf of the Appellant, Counsel who appeared for the Appellant on appeal before us, wisely did not press that attack. The evidence in support of the Respondent's claims for damages sustained on the second, third and fourth occasions when the Respondent's animals were attacked by the Appellant's dog, is based on

24
contemporaneous notes made by De Wet and by acceptable expert evidence. I am not persuaded that there is any legitimate ground on which this Court should interfere with the findings of the Trial Court which were sustained on appeal by the Provincial Division.

The Magistrate upheld the claims of the Respondent with costs. The claim of the Respondent, however, included an amount of R9 464.80 arising from the attack on the Respondent's livestock on the first occasion. Since I have held that the Respondent did not, on a balance of probabilities, establish that it was the Appellant's dog which had attacked the Respondent's animals on that occasion, it follows that that claim must be disallowed and that the damages which the Appellant must be ordered to pay to the Respondent, must be limited to those damages which were suffered by the Respondent in consequence of the attacks of the Appellant's dogs on the animals of the Respondent on the second, third and fourth

25
occasions.
The total damages claimed by the Respondent in terms of its summons as amended was R17 079.37. The damages sustained by the Respondent on the first occasion was assessed to be R9 464.80. More than 50% of the total claim for damages therefore represents the loss sustained by the Respondent on the first occasion. The Respondent has, however, not established that the Appellant's dog was responsible for these damages amounting to R9 464.80. The Appellant has therefore been substantially successful on appeal and is therefore entitled to the costs of the Appeal both in the Provincial Division and before this Court.

I accordingly order that

1.	The appeal be upheld with costs.
2.	The order made by the Magistrate is set aside and substituted by the following -

26

a) Judgment is granted in favour of

the Plaintiff in the sum of R7 614.57 (in terms of paragraphs 4.4, 5.4, and 6.4 of the Plaintiff's summons as amended).

b) The Defendant is directed to pay
the costs of the Plaintiff
including the costs consequent
upon the employment of an
advocate.

3. The Respondent is directed to pay the

Appellant's costs in the appeal to
the Cape Provincial Division of the
Supreme Court.

I MAHOMED

ACTING JUDGE OF APPEAL

NIENABER JJA)

VAN COLLER AJA) CONCUR

KUMLEBEN JA:

2

KUMLEBEN JA:

I have had the advantage of reading the judgment of my colleague Mahomed AJA. I shall refer to it, with abiding respect, as "the other judgment". I agree with the conclusion that the claims based on the second, third and fourth incidents were proved. In my view, however, it was also shown on a balance of probabilities that the appellant's dog ("the Rottweiler") was involved in the first one and that

the appellant is thus liable for this resultant damage as well. Since this is a minority judgment

the reasons for this conclusion need be no more than

tersely stated.

The relevant facts bearing upon the first

incident have been comprehensively set out in the

other judgment. Those cardinal to the question in

issue appear to me to be the following:

(i) The fact that over a period of about

3 twelve weeks before the first incident, there had been no such attacks anywhere on the respondent's farm.

(ii) The other incidents were attributable to the Rottweiler, and after its demise the marauding came to an end.

(iii) The undisputed evidence of Dyer who
said, in the words of the other judgment, "that once
a dog has started biting sheep at a particular place,
it was highly probable that it would return to repeat
such conduct". (I emphasise.)

The postulate which would non-suit the respondent as regards this incident - if as probable as any other - is that another dog, dog A, was responsible for the damage on that occasion. If that were the case, one readily concedes that as a possibility that dog could have ceased its predations for any number of reasons. That said, one
4 is still confronted with the rather striking coincidence that the Rottweiler in deciding on the place for its first excursion, should have chosen the same farm - in fact an adjoining or nearby camp on the same farm - to start its series of attacks. And would have done so within a relatively short space of time after the first incident. The interval between the first and second attack was twenty-seven days and that between the second and third incidents, which we know involved the Rottweiler, was nineteen days: not, to my mind, a noteworthy time discrepancy. On these facts in my view the natural and probable inference is that the Rottweiler, acting secundum naturam - as Dyer's evidence indicates - returned on the second occasion to the place where it had previously found its prey - and entertainment.

During argument it was suggested that more than one dog, say two, might have been involved in
5 the first incident. This consideration does not advance the appellant's case as regards the probabilities. If there were two dogs initially involved, the Rottweiler and dog A, the latter, as I have said, might for some reason have been restrained (for instance, because its owner realised what it had been up to). The Rottweiler, however, returned to the place of its former visit. If the proposition is that two other dogs, A and B, were involved, it would follow that two dogs, not one, would have for some reason ceased to act in a manner described by Dyer. Such a possibility is to my mind even more remote. Thus, in contrast to the reasoning in the other judgment, I do not consider that if the inference that more than one dog was involved is to be drawn from the number of sheep killed, it is in any way a countervailing consideration to be taken into account in deciding that the respondent had not proved its
6 case as regards this incident.

I would dismiss the appeal with costs.

M E KUMLEBEN JUDGE OF APPEAL

JOUBERT JA - Concurs

---