HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Willenburg, K. L. Articles by Knox, R. V. Search for Related Content PubMed PubMed Citation Articles by Willenburg, K. L. Articles by Knox, R. V.

Effect of boar exposure at time of insemination on factors influencing fertility in gilts¹

Department of Animal Sciences, University of Illinois, Urbana 61801

² Correspondence:
360 Animal Sciences Laboratory, 1207 W. Gregory Dr. (phone: 217- 244-5177; fax: 217-333-8286; E-mail:
rknox{at}uiuc.edu).

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

 
The effect of boar exposure during artificial insemination (AI) on semen backflow, fertilization, and embryo quality was evaluated. Gilts (~170 d) were induced into estrus with PG600, and ovulation was synchronized using hCG 72 h later. Estrus detection was initiated after PG600 and continued at 12-h intervals. At estrus, gilts were allotted to receive boar exposure (BE, n = 20) or no boar exposure (NBE, n = 20) during AI. Gilts receiving NBE were identified to be in estrus prior to AI and the boar was then removed for 1 h, whereas gilts in the BE group received 15 min of exposure during AI. Insemination occurred in crates at 12 and 24 h after onset of estrus with 3 x 10⁹ sperm/80 mL. Backflow was collected continuously with samples taken at time 0, (during AI), and at 0.25, 0.5, 0.75, 1, 2, 4, and 8 h after first and second AI. The effect of treatment was evaluated for time of insemination (min), backflow (mL), and sperm in backflow samples. Oviducts were flushed 2 d after first AI to evaluate the effect of treatment on fertilization rate, accessory sperm numbers on embryos (scored 1 to 5), and embryo quality. There was no effect of first or second AI; therefore, data were pooled. Average duration of AI was 3.7 ± 0.2 min and was not influenced by BE (P < 0.10). However, during the initial stage of AI, BE reduced the volume of semen (18.6 vs 32.4 ± 3 mL) and the number of sperm lost (0.8 vs 1.3 ± 0.15 x 10⁹ sperm) compared to NBE (P < 0.05). There was a treatment x time effect (P < 0.05) for volume of backflow. By 45 min, the BE gilts lost more volume (9.0 vs 3.6 mL) compared to the NBE group, but sperm loss did not differ. Between 1 and 8 h after AI, neither volume nor sperm loss was influenced by treatment. By 8 h, total leakage (65 vs 63 mL) and total sperm loss (1.6 x 10⁹ vs 1.8 x 10⁹ sperm) were not influenced by BE (P > 0.10). However, more accessory sperm (P < 0.01) were found on embryos for the NBE (11 sperm/embryo) compared to BE embryos (10 sperm/embryo). Despite this observation, percentages of fertilized embryos (99.5 ± 0.5 %) and number of embryos (11.5 ± 0.1) were not different (P > 0.10). In conclusion, AI in the presence of a mature boar did not affect total semen leakage, sperm loss, fertilized embryos, or embryo quality. The importance of boar exposure during insemination was evident from less leakage during insemination, but had no effect on fertility; this suggests that the elimination of boar exposure during AI may not be deleterious to reproductive performance.

Key Words: Artificial Insemination • Boars • Fertilization • Gilts • Pigs • Semen

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

 
The use of AI has increased rapidly in commercial swine production. Extensive research has been conducted to optimize the reproductive potential of AI to equal or better that of natural insemination. Despite advances, however, litter size and farrowing rate are less than optimal in most production systems (PigCHAMP, 2000). In an attempt to improve reproductive parameters in response to AI, considerable labor is performed to maintain boar contact during insemination. It has been shown that during courting, a mature boar produces derivatives of 16-androstene pheromones (Pearce and Hughes, 1987), which are found in the urine and saliva (Hughes et al., 1990). These pheromones were found to have a signaling function involved in stimulating puberty (Pearce and Hughes, 1987), and Mattioli et al. (1986) demonstrated that spraying 5-androst-16-en-3-one in front of sows for 2 s induced a rise in oxytocin. These reports suggest a role of the boar in altering the physiological response of the female. If pheromones emitted from the saliva and urine of boars influence the onset of estrus and oxytocin release, it could be hypothesized that these pheromones may also play a role in influencing sperm transport. Therefore, this study was conducted to test for the effect of boar contact during insemination on volume of semen and the number of sperm expelled from the uterus following AI, and on fertilization rate and the number of two-cell embryos. The results could provide information on the importance and necessity of providing boar stimuli during AI on the quality of insemination, sperm transport, and fertility.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

 
Experimental Design
The experiment tested for the effects of boar exposure and back pressure during artificial insemination (AI) on fluid backflow, sperm loss, fertility, and embryo quality. Fifty-eight prepubertal gilts, approximately 170 ± 10 d of age with an average weight of 110 ± 10 kg, and of mixed genetics (Landrace, Landrace-Yorkshire, or Landrace-Yorkshire-Duroc crosses), were treated with PG600 (400 IU of eCG + 200 IU of hCG; Intervet Inc., Millsboro, DE) to induce estrus. During estrus induction, gilts were group-housed in an open-front building with an operational curtain. Following estrus expression, they were moved into an environmentally controlled facility and placed in stalls for insemination. Gilts were checked for estrus twice daily by providing them with fenceline contact with a mature boar for a minimum of 15 min beginning 12 h after the injection of PG600. Seventy-two hours after PG600 treatment (0 h), all gilts were given 1,000 IU of hCG (Intervet Inc.) to induce ovulation to occur at ~40 h. After the onset of estrus (n = 40), gilts were randomly allotted by genetics, age, and weight to receive fenceline boar exposure during insemination (BE) or no boar exposure (NBE) during insemination. Once allotted to a treatment, all gilts were allowed free access to water, but limited access to feed.

Artificial Insemination
All animals received inseminations of 3 x 10⁹ sperm/80 mL at 12 and 24 h after the onset of estrus. To be included in the experiment, all gilts had to have expressed standing estrus for both the 12- and 24-h assessment. One hour prior to time of insemination, all gilts from both treatments were checked for estrus with fenceline boar exposure to ensure that all animals were still in standing estrus. All boars used in the study were housed in separate rooms (30 m away) and were brought to the gilts for estrus detection. During insemination, the animals assigned to the BE treatment were inseminated in the presence of a boar (<1 m away), with the total exposure time lasting approximately 15 min. These gilts were also fitted with a breeding saddle filled with ~9 kg of sand to mimic the back pressure stimulus. The gilts assigned to receive no boar exposure during mating after expressing standing estrus had no boar exposure for 1 h prior to AI. The NBE gilts were inseminated without boar exposure or any external back pressure applied during AI. All animals were inseminated using a spirette catheter (Minitube Inc., Verona, WI). During insemination, a 250-mL cup was held beneath the vulva to collect any semen that was lost during AI. All inseminations were timed. Timing started whenever the bottle of semen was attached to the cervix-affixed catheter and ended whenever the bottle was removed.

Backflow Collection
Backflow of semen was collected continuously for 8 h and interval measurements were obtained at insemination (0 h), 0.25, 0.5, 0.75, 1, 2, 4, and 8 h. Collection occurred using a closed system with a modified foam-tip catheter (Mintube Inc.) attached to a 50-mL conical centrifuge tube. The catheter was modified to extend 5.1 cm distal to the vulva. A 50-mL centrifuge tube excluding the lid was attached to a 2.2-cm balloon. The balloon was cut 10.2 cm from the opening and fastened to the catheter by a small rubber band. During insemination, a 250-mL cup was held beneath the vulva to collect any backflow. Backflow samples were collected and stored at 5°C, and evaluated within 24 h for volume. Number of sperm lost was determined using a hemacytometer.

In order to determine if the AI catheter collection methodology altered semen leakage, gilts (n = 7), similar in age, weight, and genetics to those used for the catheter collection method (n = 40), were evaluated for backflow following insemination with 3 x 10⁹ sperm/80 mL at 12 h after onset of estrus. For the control collection method, backflow was collected with a 250-mL cup held beneath the vulva for 2 h during and following AI. The results from both methods were compared for total leakage volume at 2 h.

Semen Collection and Handling
Whole filtered semen ejaculates were collected into a cup. Immediately following collection, semen was evaluated and extended to the proper concentration based on concentration, motility, and abnormalities. The semen was held in a water bath to maintain 37°C until the concentration was determined. After the ejaculate was extended to the appropriate concentration, it was aliquoted, sealed, and allowed to cool to room temperature. At the appropriate temperature, it was stored in a cooler programmed to maintain a temperature of 18°C. Gilts on each treatment were artificially inseminated using semen from the same boars and collections.

Embryo Collection and Analysis
At 48 h after first AI (0 h), gilts were midventrally laparotomized to evaluate ovulation rate, and to flush oviducts in order to assess fertilization rate, accessory sperm numbers on each embryo, embryo development, and quality. Embryos were collected in Beltsville embryo collection media supplemented with 1% BSA. Immediately after collection, each embryo was assessed for accessory sperm number, stage of development, and quality under a light microscope using an experienced embryo evaluator who was blind to the treatments. The stage of development was categorized on a 0 to 3 scale as follows: 0 = degenerated embryos, 1 = one cell, 2 = two cells, and 3 = four cells. The grading criteria for the number of sperm cells bound to the embryo was as follows: 1 = 1 to 5 sperm, 2 = 6 to 10 sperm, 3 = 11 to 15 sperm, 4 = 16 to 20 sperm, and 5 = greater than 20 sperm per embryo.

White Blood Cell Analysis
During the study, an increased number of nonsperm cells was observed in the backflow as time increased following insemination. In order to evaluate the cell type, a white and red blood cell differential count was performed on backflow samples from three animals per treatment for the first and second AI. The samples were submitted to the University of Illinois Veterinary Diagnostic Lab (Urbana, IL). The samples submitted to the laboratory were obtained from gilts that had backflow samples from 0, 0.25, 1, 2, 4, and 8 h after insemination.

Statistical Analysis
The data were analyzed as a completely randomized design using PROC MIXED procedures (SAS Inst., Inc., Cary, NC). In order to test for the effect of collection method on volume leaked over a 2-h period following AI, the model included method (cup vs catheter), leakage at time 0 as a covariate, and collection times following AI and their interactions for repeated measures. To test for the effect of treatment, the independent variables included treatment (BE vs NBE) and AI (first or second) as fixed variables. The continuous-response variables included volume of backflow (mL) and the number of sperm in the backflow at various intervals over an 8-h period after insemination. These variables were analyzed as a repeated measure and included treatment, time, and first or second AI. A standard variance components structure (with equal variance across repeated measurements) was specified. Ovulation rate and time to inseminate were analyzed using PROC MIXED for the effects of treatment, AI, and their interactions. The categorical responses of number of sperm bound to the embryo and embryo quality were analyzed for the effect of treatment using general linear models in SAS.

	Results

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

 
Evaluation of Collection Methodology
The catheter method of collection permitted urination, as well as the collection of all backflow, and was evaluated for its effect on rate of backflow. There was no effect of collection method on total volume leakage at the end of a 2-h period following AI (P > 0.10). The cup method resulted in a total volume of 40.9 ± 3.2 mL collected at the end of the 2-h period compared to a volume of 37.7 ± 1.4 mL for the modified catheter collection method.

For all gilts, data for backflow samples collected for each animal during the first and second inseminations were pooled because no effect of insemination period was detected (P > 0.10). During insemination, 80% of all gilts leaked semen regardless of collection method, treatment, or AI period. The volume leaked at AI time 0 ranged from 1 to 69 mL.

Effect of Treatment on Backflow
In evaluating the effect of boar exposure during insemination, there was no effect of treatment (P > 0.10) on time to inseminate (Table 1) or on backflow (Table 2). The average duration of insemination for both treatments was 3.7 min and was not influenced by boar exposure. The total volume of backflow was also not influenced by boar exposure at insemination (P > 0.10). The BE treatment lost 65 ± 2 mL, whereas the NBE treatment expelled 62 ± 2 mL. However, there was a treatment x time interaction (P < 0.05). During insemination, the NBE treatment lost 40% of the volume inseminated which was almost twice that of the BE treatment (P < 0.05). By 45 min after AI, the BE animals which lost less fluid to that point, increased volume lost (9.0 vs 3.6 mL) compared to NBE gilts. At this time (45 min), 66% of the volume inseminated was lost for both treatments.

The number of accessory sperm and embryo quality were analyzed for each individual embryo. Fertilization rates were 99.5% for the two treatments and degenerated embryos were found in 3% of the animals involved in the study. There was a difference (P < 0.01) between numbers of accessory sperm attached to the zona pellucida of d-2 embryos. The NBE group averaged 11 to 15 sperm per embryo (grade 3), whereas the BE group had a grade of 2.5 or ~8 to 10 sperm. Although significant, this only accounted for a difference ranging from 1 to 7 spermatozoa. There was no difference between the stage of development for the embryos in response to treatment (P > 0.10). The average developmental stage for each treatment was a two-cell embryo and ranged from degenerate to a four-cell embryo.

White Blood Cell Analysis
There was no difference between treatments on the average percentage of neutrophils or epithelial cells found in the backflow samples (P > 0.10). However, there was a significant interaction of first and second AI x time (P < 0.05, Table 4) on percentage of cells. During the first AI, a greater number of epithelial cells was observed in the backflow samples during the first hour. However, by 60 min post insemination, the mean percentage of neutrophils increased, and by 4 h, 98% of cells observed were neutrophils. These cells remained elevated until the last observation period at 8 h. This same immune response was prevalent for the second AI, but the neutrophil response reached its peak approximately 2 h sooner than the first AI.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

The results of this study are comparable to the findings of Steverink et al. (1998), who collected backflow with a colostomy bag for up to 2.5 h following AI for multiparous sows. They reported backflow losses equal to 70% of the total volume and 25% of the inseminated sperm. The higher values for backflow in the current study could be due to a number of different factors. Our method for backflow collection permitted urination without affecting semen collection, whereas use of a colostomy bag could not prevent urine from entering the collection bag, and would force exclusion of the contaminated samples from analysis. In our preliminary study, the system of collection was tested to validate if the catheter resulted in a greater loss of backflow compared to a natural gravitational leakage after insemination. Backflow collection with only a cup resulted in approximately the same volume as compared to the catheter. Therefore, it can be concluded that the catheter system used for collection does not facilitate an excess leakage from the uterus. Another difference that could have resulted in the increase in semen loss and sperm leakage in our study may be due to the difference in the age and size of the animals. Steverink et al. (1998) used multiparous animals, and these larger females with larger reproductive tracts could have facilitated greater retention of fluid in the tract due to gravity and position of the larger uterus as it was positioned in the peritoneal cavity. Steverink et al. (1998) reported that parity 1 animals lost an average of 5 mL more than parity 2 and older sows, and that four of the five animals that expelled 20 mL or more during insemination were parity 1 animals. Baker and Degen (1972) also reported that the size of the uterus and the number of sperm collected in backflow were correlated. Therefore, the more immature females used for this study could explain the higher percentage of backflow observed. Another likely explanation for the increased difference could be related to the fact that backflow was collected for 8 h after insemination compared to the 2.5-h collection time of Steverink et al. (1998).

During the study, an increased number of cells (not sperm) were observed in the backflow samples, which led us to perform a differential cell count. A greater number of epithelial cells was observed in the backflow samples within the first hour following the first insemination. By 1 h, the percentage of neutrophils increased, and by 4 h after insemination, almost 100% of cells counted were neutrophils. This deviates from the findings of Rozeboom et al. (1998), who observed that the inflammatory response reaches its magnitude between 6 and 12 h after AI. Perhaps one explanation is that their measurements did not start until 6 h after insemination, whereas in the present experiment, the first measurement occurred 15 min after insemination. A similar trend for increased neutrophils was observed for the second AI, except that the increase in neutrophils reached its peak by 2 h, compared to 4 h following the first AI. The reason for this increase in neutrophils is due to the finding that spermatozoa in the uterus causes an influx of polymorphonuclear neutrophilic granulocytes via complement to the site of inflammation (Troedsson et al., 1998). The difference in immune response between the first and second AI could be due the effect of the induced inflammatory response resulting from the first insemination.

Fertilization rate was unaffected by inseminating in the presence of a boar. Both treatments had nearly a 100% fertilization rate. Polge (1978) reported an 85% embryo recovery rate and a 95% fertilization rate for gilts inseminated with 80 to 100 mL of fresh undiluted semen on the second day of estrus or 24 h after the injection of hCG. These results are similar to the current experiment. The high fertilization rate could have resulted from controlling the time of ovulation and utilizing a double insemination. Differences may have been observed had hCG not been administered and a single insemination used. Our data indicate that there was no difference in backflow rates or sperm loss between the first and second insemination. During estrus, the uterus is a highly volatile environment because it is under constant hormonal control. Zerobin and Sporri (1972) described variation in uterine contractions from day to day. Contractions on the second day of estrus were more favorable to sperm transport than on d 1 or 3. Based on this report, it would seem that the rate of backflow would differ as ovulation approached, but that did not appear to be the case in the current study.

There was a significant difference in the number of accessory sperm between the two treatments, but the NBE group had less than 10 more sperm for each embryo. DeJarnette et al. (1992) reported that large numbers of accessory sperm positively affect embryo quality. However, in their study, embryos were collected on d 6 and evaluated at the morula and blastocyst stage. Steverink et al. (1997) stated that accessory sperm numbers were related to the insemination-ovulation interval, and when inseminating 3 x 10⁹ sperm 12 to 24 h before ovulation, a mean accessory number of 17 was observed, comparable to the accessory sperm number in the present study. These authors stated that knowing the insemination-to-ovulation interval is necessary for interpreting the relationship between accessory sperm number and embryo quality. Although the time of ovulation was not known in our study, it can be assumed that ovulation occurred 40 h after the injection of hCG (Dziuk and Baker, 1962). Based on this assumption, all inseminations occurred approximately 12 h prior to ovulation.

Inseminating gilts in the presence of a mature boar did not affect overall leakage, fertilization rate, or embryo quality, even though a difference in accessory sperm numbers on the embryos was observed to be higher with no boar exposure. This apparently had no impact on our assessment of early embryo quality. The positive effect of BE was observed during insemination when the gilts displayed a rigid state of standing estrus, allowing for an easier insemination and less fluid loss at time 0. However, the total volume of backflow over the 8-h period was proportional to the volume expelled during insemination at time 0. A greater loss of volume during insemination results in a lesser amount of volume over time and vice versa. The proportion of spermatozoa lost was also directly related to the volume of semen expelled at the corresponding time point and is in agreement with the conclusions of Baker et al. (1968) regarding patterns of sperm loss in gilts. This leads us to the conclusion that boar exposure during insemination may not be necessary when conditions for fertility are optimal. In situations where an insemination must occur without a boar, no reduction in fertility would be expected. This could be useful for time and labor savings associated with boar movement. In this study, we conclude that optimal conditions for fertility may include twice-daily estrus detection, insemination times that average 3 to 4 min/female, a sperm concentration of 3 billion fertile sperm, and a double insemination that provides that last insemination at 12 h before ovulation. It is not known whether boar contact during AI would be more critical in situations when estrus detection, time of insemination, and semen fertility are less than optimal.

	Implications

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

 
Boar exposure and back pressure during insemination did not produce an overall advantage in lowering sperm loss, semen backflow, ovulation rate, fertilization rate, or on the number and quality of embryos compared to performing artificial insemination in the absence of a boar. However, the effect of boar exposure was observed during insemination and allowed for an easier artificial insemination, with a slightly reduced time to inseminate, and a lower volume of sperm and volume lost. It appears that when conditions for fertility are optimal, inseminating in the presence of a boar may not be critical. Optimal conditions for fertility may include a sperm concentration of 3 billion fertile sperm, a double insemination, and twice-daily estrus detection. The full potential of a boar effect during AI may be more critical in situations when estrus detection, time of insemination, and semen fertility are less than optimal.

	Footnotes

 
¹ This research was supported in part by the Illinois Council on Food and Agricultural Research (C-FAR) and the Dept. of Animal Science, Univ. of Illinois. The animal care and use committee of the Univ. of Illinois approved the protocol for use of animals in this experiment.

Received for publication May 30, 2002. Accepted for publication August 22, 2002.

	Literature Cited

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

 


Baker, R. D., and A. A. Degen. 1972. Transport of live and dead boar spermatozoa within the reproductive tract of gilts. J. Reprod. Fertil. 28:369–377.[Abstract/Free Full Text]

Baker, R. D., P. J. Dziuk, and H. W. Norton. 1968. Effect of volume of semen, number of sperm and drugs in transport of sperm in artificially inseminated gilts. J. Anim. Sci. 27:88–93.

DeJarnette, J. M., R. G. Saacke, J. Bame, and C. J. Vogler. 1992. Accessory sperm: their importance to fertility and embryo quality, and attempts to alter their numbers in artificially inseminated cattle. J. Anim. Sci. 70:484–491.[Abstract]

Dziuk, P. J., and R. D. Baker. 1962. Induction and control of ovulation in swine. J. Anim. Sci. 21:697–699.[Abstract/Free Full Text]

Hughes, P. E., G. P. Pearce, and A. M. Patterson. 1990. Mechanisms mediating the stimulatory effects of the boar on gilt reproduction. Control of Pig Reproduction III. J. Reprod. Fertil. (Suppl.)40:323–341.

Mattioli, M., G. Galeati, F. Conte, and E. Seren. 1986. Effect of 5(-androst-16-en-3-one on oxytocin release in oestrus sows. Theriogenology 25:399–403.

Pearce, G. P. and P. E. Hughes. 1987. The influence of boar-component stimuli on puberty attainment in the gilt. Anim. Prod. 44:293–302.

PigCHAMP. 2000. Available: http://showcase.netins.net/web/swinedata/reports.html. Accessed April 12, 2001.

Polge, C. 1978. Fertilization in the pig and horse. J. Reprod. Fertil. 54:461–470.[Abstract/Free Full Text]

Rozeboom, K. J., M. H. T. Troedsson, and B. G. Crabo. 1998. Characterization of uterine leukocytes in gilts after artificial insemination. J. Reprod. Fertil. 114:195–199.[Abstract/Free Full Text]

Soede, N. M. 1993. Boar stimuli around insemination affect reproductive processes in pigs: A review. Anim. Reprod. Sci. 32:107–125.

Steverink, D. W. B., N. M. Soede, E. G. Bouwman, and B. Kemp. 1997. Influence of insemination-ovulation interval and sperm cell dose on fertilization in sows. J. Reprod. Fertil. 111:165–171.[Abstract/Free Full Text]

Steverink, D. W. B., N. M. Soede, E. G. Bouwman, and B. Kemp. 1998. Semen backflow and its effect on fertilization results in sows. Anim. Reprod. Sci. 54:109–119.[Medline]

Troedsson, M. H. T., I. K. M. Liu, and B. G. Crabo. 1998. Sperm transport and survival in the female genital tract. Theriogenology 49:905–915.[Medline]

Zerobin, K., and H. Sporri. 1972. Motility of the bovine and porcine uterus and fallopian tube. Adv. Vet. Sci. Comp. Med. 16:303–354[Medline]

This article has been cited by other articles:

				K. L. Willenburg, G. M. Miller, S. L. Rodriguez-Zas, and R. V. Knox Influence of hormone supplementation to extended semen on artificial insemination, uterine contractions, establishment of a sperm reservoir, and fertility in swine J Anim Sci, April 1, 2003; 81(4): 821 - 829. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Willenburg, K. L. Articles by Knox, R. V. Search for Related Content PubMed PubMed Citation Articles by Willenburg, K. L. Articles by Knox, R. V.